{
  "name": "AI Release Tracker — model releases",
  "description": "Every major AI model release tracked since the launch of ChatGPT on November 30 2022, with release dates, published benchmark scores, parameter counts, context windows and licensing tier.",
  "source": "https://aireleasetracker.com",
  "license": "https://aireleasetracker.com/terms",
  "citation": "AI Release Tracker (https://aireleasetracker.com). Benchmark scores are the figures published by the releasing lab at launch.",
  "lastModified": "2026-08-14T00:00:00.000Z",
  "counts": {
    "models": 229,
    "companies": 11
  },
  "benchmarks": [
    {
      "key": "bullshitbench-v2",
      "name": "BullshitBench v2",
      "variant": null,
      "category": "robustness",
      "unit": "%",
      "higherIsBetter": true,
      "description": "detecting and pushing back on nonsense prompts",
      "url": "https://aireleasetracker.com/benchmark/bullshitbench-v2"
    },
    {
      "key": "grayswan-ipi-k1",
      "name": "Gray Swan IPI",
      "variant": "k = 1",
      "category": "robustness",
      "unit": "%",
      "higherIsBetter": false,
      "description": "indirect prompt injection attack success rate",
      "url": "https://aireleasetracker.com/benchmark/grayswan-ipi-k1"
    },
    {
      "key": "grayswan-ipi-k10",
      "name": "Gray Swan IPI",
      "variant": "k = 10",
      "category": "robustness",
      "unit": "%",
      "higherIsBetter": false,
      "description": "indirect prompt injection attack success rate",
      "url": "https://aireleasetracker.com/benchmark/grayswan-ipi-k10"
    },
    {
      "key": "grayswan-ipi-k15",
      "name": "Gray Swan IPI",
      "variant": "k = 15",
      "category": "robustness",
      "unit": "%",
      "higherIsBetter": false,
      "description": "indirect prompt injection attack success rate",
      "url": "https://aireleasetracker.com/benchmark/grayswan-ipi-k15"
    },
    {
      "key": "swe-bench-pro",
      "name": "SWE-Bench Pro",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "long-horizon, real-world software engineering",
      "url": "https://aireleasetracker.com/benchmark/swe-bench-pro"
    },
    {
      "key": "swe-bench-verified",
      "name": "SWE-Bench Verified",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "real-world software engineering",
      "url": "https://aireleasetracker.com/benchmark/swe-bench-verified"
    },
    {
      "key": "swe-bench-multilingual",
      "name": "SWE-Bench Multilingual",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "software engineering across many languages",
      "url": "https://aireleasetracker.com/benchmark/swe-bench-multilingual"
    },
    {
      "key": "cursorbench-v3.2",
      "name": "CursorBench v3.2",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Cursor agentic coding eval (v3.2 task set)",
      "url": "https://aireleasetracker.com/benchmark/cursorbench-v3.2"
    },
    {
      "key": "cursorbench-v3.1",
      "name": "CursorBench v3.1",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Cursor agentic coding eval (harder tasks)",
      "url": "https://aireleasetracker.com/benchmark/cursorbench-v3.1"
    },
    {
      "key": "deepswe-1.1",
      "name": "DeepSWE 1.1",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Artificial Analysis agentic software-engineering eval (v1.1)",
      "url": "https://aireleasetracker.com/benchmark/deepswe-1.1"
    },
    {
      "key": "deepswe-1.0",
      "name": "DeepSWE 1.0",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Artificial Analysis agentic software-engineering eval",
      "url": "https://aireleasetracker.com/benchmark/deepswe-1.0"
    },
    {
      "key": "frontiercode-1.1",
      "name": "FrontierCode v1.1 (Main)",
      "variant": "main split",
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "frontier-difficulty agentic coding tasks (v1.1, main split)",
      "url": "https://aireleasetracker.com/benchmark/frontiercode-1.1"
    },
    {
      "key": "frontiercode-1.1-extended",
      "name": "FrontierCode v1.1 (Extended)",
      "variant": "extended split",
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "frontier-difficulty agentic coding tasks (v1.1, extended split)",
      "url": "https://aireleasetracker.com/benchmark/frontiercode-1.1-extended"
    },
    {
      "key": "apex-swe",
      "name": "APEX-SWE",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert-level software-engineering tasks (AI Productivity Index)",
      "url": "https://aireleasetracker.com/benchmark/apex-swe"
    },
    {
      "key": "mle-bench",
      "name": "MLE-Bench",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "machine-learning engineering on real Kaggle competitions",
      "url": "https://aireleasetracker.com/benchmark/mle-bench"
    },
    {
      "key": "paperbench",
      "name": "PaperBench",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "reproducing the results of an ML research paper end to end",
      "url": "https://aireleasetracker.com/benchmark/paperbench"
    },
    {
      "key": "nextjs-evals",
      "name": "Next.js Evals",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Next.js code generation & migration tasks",
      "url": "https://aireleasetracker.com/benchmark/nextjs-evals"
    },
    {
      "key": "supabase-evals",
      "name": "Supabase Evals",
      "variant": "with skills",
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "building and debugging real Supabase projects as a coding agent",
      "url": "https://aireleasetracker.com/benchmark/supabase-evals"
    },
    {
      "key": "supabase-evals-no-skills",
      "name": "Supabase Evals",
      "variant": "no skills",
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "building and debugging real Supabase projects as a coding agent",
      "url": "https://aireleasetracker.com/benchmark/supabase-evals-no-skills"
    },
    {
      "key": "nl2repo-bench",
      "name": "NL2Repo-Bench",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "generating repository-level code from natural-language requirements",
      "url": "https://aireleasetracker.com/benchmark/nl2repo-bench"
    },
    {
      "key": "qwenswebench",
      "name": "QwenSWEBench",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "Qwen's in-house benchmark for software-engineering capabilities",
      "url": "https://aireleasetracker.com/benchmark/qwenswebench"
    },
    {
      "key": "livecodebench",
      "name": "LiveCodeBench",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "contamination-free code generation on freshly published problems",
      "url": "https://aireleasetracker.com/benchmark/livecodebench"
    },
    {
      "key": "humaneval",
      "name": "HumanEval",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "writing a single Python function from its docstring — the standard coding benchmark of 2021–2024, retired in favour of SWE-Bench-style tests on real repositories",
      "url": "https://aireleasetracker.com/benchmark/humaneval"
    },
    {
      "key": "frontier-bench-v0.1",
      "name": "Frontier-Bench v0.1",
      "variant": null,
      "category": "terminal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "diverse, difficult agentic computer-work tasks (successor to Terminal-Bench)",
      "url": "https://aireleasetracker.com/benchmark/frontier-bench-v0.1"
    },
    {
      "key": "terminal-bench-3.0",
      "name": "Terminal-Bench 3.0",
      "variant": null,
      "category": "terminal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "command-line task completion (v3.0, much harder task set)",
      "url": "https://aireleasetracker.com/benchmark/terminal-bench-3.0"
    },
    {
      "key": "terminal-bench-2.1",
      "name": "Terminal-Bench 2.1",
      "variant": null,
      "category": "terminal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "command-line task completion",
      "url": "https://aireleasetracker.com/benchmark/terminal-bench-2.1"
    },
    {
      "key": "terminal-bench-2.0",
      "name": "Terminal-Bench 2.0",
      "variant": null,
      "category": "terminal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "command-line task completion (v2.0)",
      "url": "https://aireleasetracker.com/benchmark/terminal-bench-2.0"
    },
    {
      "key": "expert-swe",
      "name": "Expert-SWE (Internal)",
      "variant": null,
      "category": "coding",
      "unit": "%",
      "higherIsBetter": true,
      "description": "OpenAI's internal software-engineering eval",
      "url": "https://aireleasetracker.com/benchmark/expert-swe"
    },
    {
      "key": "apex-agents",
      "name": "APEX-Agents",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert-level agentic work tasks (AI Productivity Index)",
      "url": "https://aireleasetracker.com/benchmark/apex-agents"
    },
    {
      "key": "mcp-atlas",
      "name": "MCP Atlas",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "multi-step workflows using MCP",
      "url": "https://aireleasetracker.com/benchmark/mcp-atlas"
    },
    {
      "key": "jobbench",
      "name": "JobBench",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "professional workplace tool-use tasks",
      "url": "https://aireleasetracker.com/benchmark/jobbench"
    },
    {
      "key": "coworkbench",
      "name": "CoWorkBench",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "long-horizon productivity tasks across professional domains",
      "url": "https://aireleasetracker.com/benchmark/coworkbench"
    },
    {
      "key": "toolathlon-verified",
      "name": "Toolathlon-Verified",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "verified real-world personal tool use",
      "url": "https://aireleasetracker.com/benchmark/toolathlon-verified"
    },
    {
      "key": "toolathlon",
      "name": "Toolathlon",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "real-world general tool use",
      "url": "https://aireleasetracker.com/benchmark/toolathlon"
    },
    {
      "key": "bu-bench",
      "name": "BU Bench",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "browser agent task completion (Browser Use)",
      "url": "https://aireleasetracker.com/benchmark/bu-bench"
    },
    {
      "key": "browsecomp",
      "name": "BrowseComp",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "agentic web browsing and research",
      "url": "https://aireleasetracker.com/benchmark/browsecomp"
    },
    {
      "key": "cybergym",
      "name": "CyberGym",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "cybersecurity agentic tasks",
      "url": "https://aireleasetracker.com/benchmark/cybergym"
    },
    {
      "key": "exploitbench",
      "name": "ExploitBench",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "exploit development on known V8 vulnerabilities (CMU capability-ladder eval)",
      "url": "https://aireleasetracker.com/benchmark/exploitbench"
    },
    {
      "key": "exploitgym-6h",
      "name": "ExploitGym",
      "variant": "6-hour budget",
      "category": "agentic",
      "unit": "score",
      "higherIsBetter": true,
      "description": "end-to-end exploits built from known bugs (count of 898 instances solved)",
      "url": "https://aireleasetracker.com/benchmark/exploitgym-6h"
    },
    {
      "key": "exploitgym-2h",
      "name": "ExploitGym",
      "variant": "2-hour budget",
      "category": "agentic",
      "unit": "score",
      "higherIsBetter": true,
      "description": "end-to-end exploits built from known bugs (count of 898 instances solved)",
      "url": "https://aireleasetracker.com/benchmark/exploitgym-2h"
    },
    {
      "key": "hle-no-tools",
      "name": "Humanity's Last Exam",
      "variant": "no tools",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert-level questions across disciplines",
      "url": "https://aireleasetracker.com/benchmark/hle-no-tools"
    },
    {
      "key": "hle-with-tools",
      "name": "Humanity's Last Exam",
      "variant": "with tools",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert-level questions across disciplines",
      "url": "https://aireleasetracker.com/benchmark/hle-with-tools"
    },
    {
      "key": "hle-verified",
      "name": "Humanity's Last Exam (Verified)",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert-level questions across disciplines, on the verified question set",
      "url": "https://aireleasetracker.com/benchmark/hle-verified"
    },
    {
      "key": "arc-agi-3",
      "name": "ARC-AGI-3",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "novel problem-solving in interactive environments",
      "url": "https://aireleasetracker.com/benchmark/arc-agi-3"
    },
    {
      "key": "arc-agi-2",
      "name": "ARC-AGI-2",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "abstract reasoning puzzles",
      "url": "https://aireleasetracker.com/benchmark/arc-agi-2"
    },
    {
      "key": "frontiermath-t13",
      "name": "FrontierMath",
      "variant": "Tier 1–3",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "research-level mathematics",
      "url": "https://aireleasetracker.com/benchmark/frontiermath-t13"
    },
    {
      "key": "frontiermath-t4",
      "name": "FrontierMath",
      "variant": "Tier 4",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "research-level mathematics",
      "url": "https://aireleasetracker.com/benchmark/frontiermath-t4"
    },
    {
      "key": "biomysterybench-hard",
      "name": "BioMysteryBench",
      "variant": "hard",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "solving open biological research mysteries",
      "url": "https://aireleasetracker.com/benchmark/biomysterybench-hard"
    },
    {
      "key": "biomysterybench-human-solved",
      "name": "BioMysteryBench",
      "variant": "human solved",
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "solving open biological research mysteries",
      "url": "https://aireleasetracker.com/benchmark/biomysterybench-human-solved"
    },
    {
      "key": "labbench2",
      "name": "LAB-Bench 2",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "real-world biology research tasks",
      "url": "https://aireleasetracker.com/benchmark/labbench2"
    },
    {
      "key": "gpqa",
      "name": "GPQA Diamond",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "graduate-level science reasoning",
      "url": "https://aireleasetracker.com/benchmark/gpqa"
    },
    {
      "key": "ifbench",
      "name": "IFBench",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "following detailed instructions and constraints",
      "url": "https://aireleasetracker.com/benchmark/ifbench"
    },
    {
      "key": "mmlu",
      "name": "MMLU",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "multiple-choice exam questions across 57 academic subjects",
      "url": "https://aireleasetracker.com/benchmark/mmlu"
    },
    {
      "key": "gsm8k",
      "name": "GSM8K",
      "variant": null,
      "category": "reasoning",
      "unit": "%",
      "higherIsBetter": true,
      "description": "grade-school maths word problems needing a few steps of arithmetic reasoning",
      "url": "https://aireleasetracker.com/benchmark/gsm8k"
    },
    {
      "key": "osworld-2.0",
      "name": "OSWorld 2.0",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "operating a real computer to finish tasks (v2.0)",
      "url": "https://aireleasetracker.com/benchmark/osworld-2.0"
    },
    {
      "key": "osworld-verified",
      "name": "OSWorld-Verified",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "operating a real computer to finish tasks",
      "url": "https://aireleasetracker.com/benchmark/osworld-verified"
    },
    {
      "key": "agents-last-exam",
      "name": "Agent's Last Exam",
      "variant": "pass@1",
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "pass rate on multimodal desktop and OS agent tasks",
      "url": "https://aireleasetracker.com/benchmark/agents-last-exam"
    },
    {
      "key": "agents-last-exam-score",
      "name": "Agent's Last Exam",
      "variant": "score",
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "graded score on multimodal desktop and OS agent tasks",
      "url": "https://aireleasetracker.com/benchmark/agents-last-exam-score"
    },
    {
      "key": "automationbench",
      "name": "AutomationBench",
      "variant": null,
      "category": "agentic",
      "unit": "%",
      "higherIsBetter": true,
      "description": "end-to-end business workflow automation",
      "url": "https://aireleasetracker.com/benchmark/automationbench"
    },
    {
      "key": "finance-agent-v2",
      "name": "Finance Agent v2",
      "variant": null,
      "category": "finance",
      "unit": "%",
      "higherIsBetter": true,
      "description": "financial analysis tasks",
      "url": "https://aireleasetracker.com/benchmark/finance-agent-v2"
    },
    {
      "key": "harvey-legal-agent-heldout",
      "name": "Harvey's Legal Agent Benchmark (Held-out)",
      "variant": null,
      "category": "legal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "completing legal work on a held-out task set",
      "url": "https://aireleasetracker.com/benchmark/harvey-legal-agent-heldout"
    },
    {
      "key": "harvey-legal-agent",
      "name": "Harvey's Legal Agent Benchmark",
      "variant": null,
      "category": "legal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "completing legal work with documents, spreadsheets, presentations, and file-system tools",
      "url": "https://aireleasetracker.com/benchmark/harvey-legal-agent"
    },
    {
      "key": "taxeval-v2",
      "name": "TaxEval v2",
      "variant": null,
      "category": "legal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "answering tax questions (Vals AI proprietary set)",
      "url": "https://aireleasetracker.com/benchmark/taxeval-v2"
    },
    {
      "key": "healthbench-professional",
      "name": "HealthBench Professional",
      "variant": null,
      "category": "healthcare",
      "unit": "%",
      "higherIsBetter": true,
      "description": "realistic health conversations graded against physician rubrics",
      "url": "https://aireleasetracker.com/benchmark/healthbench-professional"
    },
    {
      "key": "medscribe",
      "name": "MedScribe",
      "variant": null,
      "category": "healthcare",
      "unit": "%",
      "higherIsBetter": true,
      "description": "supporting doctors with administrative work (Vals AI proprietary set)",
      "url": "https://aireleasetracker.com/benchmark/medscribe"
    },
    {
      "key": "aa-intelligence-index",
      "name": "AA Intelligence Index",
      "variant": null,
      "category": "knowledge",
      "unit": "score",
      "higherIsBetter": true,
      "description": "Artificial Analysis composite intelligence index across evals",
      "url": "https://aireleasetracker.com/benchmark/aa-intelligence-index"
    },
    {
      "key": "gdpval-aa",
      "name": "GDPval-AA",
      "variant": null,
      "category": "knowledge",
      "unit": "score",
      "higherIsBetter": true,
      "description": "economically valuable knowledge work",
      "url": "https://aireleasetracker.com/benchmark/gdpval-aa"
    },
    {
      "key": "gdpval-aa-v2",
      "name": "GDPval-AA v2",
      "variant": null,
      "category": "knowledge",
      "unit": "score",
      "higherIsBetter": true,
      "description": "economically valuable knowledge work (v2, re-based Elo)",
      "url": "https://aireleasetracker.com/benchmark/gdpval-aa-v2"
    },
    {
      "key": "aa-briefcase",
      "name": "AA-Briefcase",
      "variant": null,
      "category": "knowledge",
      "unit": "score",
      "higherIsBetter": true,
      "description": "Artificial Analysis agentic office-work eval (Elo)",
      "url": "https://aireleasetracker.com/benchmark/aa-briefcase"
    },
    {
      "key": "gdpval-winrate",
      "name": "GDPval (win/tie rate)",
      "variant": null,
      "category": "knowledge",
      "unit": "%",
      "higherIsBetter": true,
      "description": "GDPval win-or-tie rate vs human experts",
      "url": "https://aireleasetracker.com/benchmark/gdpval-winrate"
    },
    {
      "key": "charxiv-reasoning",
      "name": "CharXiv Reasoning",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "information synthesis from complex charts",
      "url": "https://aireleasetracker.com/benchmark/charxiv-reasoning"
    },
    {
      "key": "gdp-pdf",
      "name": "GDP.PDF",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "expert comprehension of PDF documents",
      "url": "https://aireleasetracker.com/benchmark/gdp-pdf"
    },
    {
      "key": "lvbench",
      "name": "LVBench",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "understanding hour-long videos",
      "url": "https://aireleasetracker.com/benchmark/lvbench"
    },
    {
      "key": "babyvision",
      "name": "BabyVision",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "core visual reasoning",
      "url": "https://aireleasetracker.com/benchmark/babyvision"
    },
    {
      "key": "mmmu-pro",
      "name": "MMMU-Pro",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "multimodal understanding and reasoning",
      "url": "https://aireleasetracker.com/benchmark/mmmu-pro"
    },
    {
      "key": "mmmu",
      "name": "MMMU",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "multimodal understanding",
      "url": "https://aireleasetracker.com/benchmark/mmmu"
    },
    {
      "key": "blueprint-bench-2",
      "name": "Blueprint-Bench 2",
      "variant": null,
      "category": "multimodal",
      "unit": "%",
      "higherIsBetter": true,
      "description": "agentic spatial reasoning",
      "url": "https://aireleasetracker.com/benchmark/blueprint-bench-2"
    },
    {
      "key": "mrcr-v2-128k",
      "name": "MRCR v2 (8-needle)",
      "variant": "128k average",
      "category": "long-context",
      "unit": "%",
      "higherIsBetter": true,
      "description": "long-context retrieval",
      "url": "https://aireleasetracker.com/benchmark/mrcr-v2-128k"
    },
    {
      "key": "mrcr-v2-1m",
      "name": "MRCR v2 (8-needle)",
      "variant": "1M pointwise",
      "category": "long-context",
      "unit": "%",
      "higherIsBetter": true,
      "description": "long-context retrieval",
      "url": "https://aireleasetracker.com/benchmark/mrcr-v2-1m"
    },
    {
      "key": "arena-text-elo",
      "name": "Arena Elo (Text)",
      "variant": null,
      "category": "preference",
      "unit": "elo",
      "higherIsBetter": true,
      "description": "human head-to-head preference votes on arena.ai (text & chat)",
      "url": "https://aireleasetracker.com/benchmark/arena-text-elo"
    },
    {
      "key": "arena-code-elo",
      "name": "Arena Elo (Code)",
      "variant": null,
      "category": "preference",
      "unit": "elo",
      "higherIsBetter": true,
      "description": "human head-to-head preference votes on arena.ai (code generation)",
      "url": "https://aireleasetracker.com/benchmark/arena-code-elo"
    }
  ],
  "companies": [
    {
      "key": "openai",
      "name": "OpenAI",
      "description": "Creators of ChatGPT and the GPT series of models. Pioneered large-scale language model research.",
      "website": "https://openai.com",
      "sameAs": [
        "https://en.wikipedia.org/wiki/OpenAI",
        "https://x.com/OpenAI",
        "https://github.com/openai"
      ],
      "releases": 43,
      "url": "https://aireleasetracker.com/company/openai"
    },
    {
      "key": "anthropic",
      "name": "Anthropic",
      "description": "AI safety company building the Claude family of models. Founded in 2021 by former OpenAI researchers.",
      "website": "https://www.anthropic.com",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Anthropic",
        "https://x.com/AnthropicAI",
        "https://github.com/anthropics"
      ],
      "releases": 24,
      "url": "https://aireleasetracker.com/company/anthropic"
    },
    {
      "key": "google",
      "name": "Google",
      "description": "Builds the Gemini family of models through Google DeepMind. Integrates AI across Google products.",
      "website": "https://deepmind.google",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Google_DeepMind",
        "https://x.com/GoogleDeepMind"
      ],
      "releases": 28,
      "url": "https://aireleasetracker.com/company/google"
    },
    {
      "key": "meta",
      "name": "Meta",
      "description": "Develops the open-weight Llama series of models. Committed to open-source AI research.",
      "website": "https://ai.meta.com",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Meta_AI",
        "https://x.com/AIatMeta",
        "https://github.com/meta-llama"
      ],
      "releases": 14,
      "url": "https://aireleasetracker.com/company/meta"
    },
    {
      "key": "xai",
      "name": "SpaceXAI",
      "description": "Elon Musk's AI company building the Grok series of models. Founded in 2023 as xAI, now part of SpaceX.",
      "website": "https://x.ai",
      "sameAs": [
        "https://en.wikipedia.org/wiki/XAI_(company)",
        "https://x.com/xai",
        "https://github.com/xai-org"
      ],
      "releases": 16,
      "url": "https://aireleasetracker.com/company/xai"
    },
    {
      "key": "deepseek",
      "name": "DeepSeek",
      "description": "Chinese AI lab known for efficient, open-weight models. Gained attention for strong performance at lower cost.",
      "website": "https://www.deepseek.com",
      "sameAs": [
        "https://en.wikipedia.org/wiki/DeepSeek",
        "https://x.com/deepseek_ai",
        "https://github.com/deepseek-ai"
      ],
      "releases": 22,
      "url": "https://aireleasetracker.com/company/deepseek"
    },
    {
      "key": "mistral",
      "name": "Mistral",
      "description": "French AI company building open and commercial models. Founded in 2023 by former Meta and DeepMind researchers.",
      "website": "https://mistral.ai",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Mistral_AI",
        "https://x.com/MistralAI",
        "https://github.com/mistralai"
      ],
      "releases": 32,
      "url": "https://aireleasetracker.com/company/mistral"
    },
    {
      "key": "moonshot",
      "name": "Moonshot AI",
      "description": "Chinese AI company building the Kimi family of models. Founded in 2023, known for long-context and efficient MoE architectures.",
      "website": "https://www.moonshot.ai",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Moonshot_AI",
        "https://x.com/Kimi_Moonshot",
        "https://github.com/MoonshotAI"
      ],
      "releases": 12,
      "url": "https://aireleasetracker.com/company/moonshot"
    },
    {
      "key": "cursor",
      "name": "Cursor",
      "description": null,
      "website": "https://cursor.com",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Cursor_(code_editor)",
        "https://x.com/cursor_ai"
      ],
      "releases": 4,
      "url": "https://aireleasetracker.com/company/cursor"
    },
    {
      "key": "zai",
      "name": "Z.ai",
      "description": "Chinese AI lab spun out of Tsinghua University (formerly Zhipu AI), building the open-weight GLM family. Rebranded internationally as Z.ai in 2025.",
      "website": "https://z.ai",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Zhipu_AI",
        "https://x.com/Zai_org",
        "https://github.com/zai-org"
      ],
      "releases": 12,
      "url": "https://aireleasetracker.com/company/zai"
    },
    {
      "key": "qwen",
      "name": "Qwen",
      "description": "Alibaba's AI lab, building the Qwen family. The most prolific publisher of open-weight models of any major lab, alongside a proprietary Max and Plus tier sold through Alibaba Cloud.",
      "website": "https://qwen.ai",
      "sameAs": [
        "https://en.wikipedia.org/wiki/Qwen",
        "https://x.com/Alibaba_Qwen",
        "https://github.com/QwenLM",
        "https://huggingface.co/Qwen"
      ],
      "releases": 22,
      "url": "https://aireleasetracker.com/company/qwen"
    }
  ],
  "models": [
    {
      "name": "GLM-5.3",
      "slug": "glm-5.3",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2026-08-14",
      "releaseDateLabel": "Aug 14 2026",
      "access": "proprietary",
      "type": null,
      "parameters": "743B",
      "contextWindow": null,
      "benchmarks": {
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 66.9,
          "unit": "%"
        },
        "terminal-bench-3.0": {
          "name": "Terminal-Bench 3.0",
          "value": 28.3,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 84.5,
          "unit": "%"
        },
        "exploitbench": {
          "name": "ExploitBench",
          "value": 54.4,
          "unit": "%"
        },
        "exploitgym-6h": {
          "name": "ExploitGym (6-hour budget)",
          "value": 130,
          "unit": "score"
        },
        "exploitgym-2h": {
          "name": "ExploitGym (2-hour budget)",
          "value": 105,
          "unit": "score"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 62.5,
          "unit": "%"
        },
        "agents-last-exam": {
          "name": "Agent's Last Exam (pass@1)",
          "value": 28.5,
          "unit": "%"
        },
        "automationbench": {
          "name": "AutomationBench",
          "value": 48.2,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1769,
          "unit": "score"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-5.3",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-5.3"
    },
    {
      "name": "Qwen3.8-27B",
      "slug": "qwen3.8-27b",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-08-14",
      "releaseDateLabel": "Aug 14 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "27B",
      "contextWindow": "262k",
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 61.7,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 42.2,
          "unit": "%"
        },
        "nl2repo-bench": {
          "name": "NL2Repo-Bench",
          "value": 42.3,
          "unit": "%"
        },
        "qwenswebench": {
          "name": "QwenSWEBench",
          "value": 79,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 90.3,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 73,
          "unit": "%"
        },
        "jobbench": {
          "name": "JobBench",
          "value": 33.4,
          "unit": "%"
        },
        "coworkbench": {
          "name": "CoWorkBench",
          "value": 70.7,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 30.8,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 89.2,
          "unit": "%"
        },
        "ifbench": {
          "name": "IFBench",
          "value": 79.5,
          "unit": "%"
        },
        "agents-last-exam": {
          "name": "Agent's Last Exam (pass@1)",
          "value": 20.4,
          "unit": "%"
        },
        "agents-last-exam-score": {
          "name": "Agent's Last Exam (score)",
          "value": 42.9,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.8-27b",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.8-27b"
    },
    {
      "name": "Gemini 3.7 Flash",
      "slug": "gemini-3.7-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-08-13",
      "releaseDateLabel": "Aug 13 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 65.3,
          "unit": "%"
        },
        "frontiercode-1.1": {
          "name": "FrontierCode v1.1 (Main) (main split)",
          "value": 43.6,
          "unit": "%"
        },
        "terminal-bench-3.0": {
          "name": "Terminal-Bench 3.0",
          "value": 14.9,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 85.8,
          "unit": "%"
        },
        "hle-verified": {
          "name": "Humanity's Last Exam (Verified)",
          "value": 53.6,
          "unit": "%"
        },
        "biomysterybench-hard": {
          "name": "BioMysteryBench (hard)",
          "value": 43.5,
          "unit": "%"
        },
        "biomysterybench-human-solved": {
          "name": "BioMysteryBench (human solved)",
          "value": 87.1,
          "unit": "%"
        },
        "labbench2": {
          "name": "LAB-Bench 2",
          "value": 82.1,
          "unit": "%"
        },
        "osworld-2.0": {
          "name": "OSWorld 2.0",
          "value": 38.1,
          "unit": "%"
        },
        "agents-last-exam": {
          "name": "Agent's Last Exam (pass@1)",
          "value": 26.3,
          "unit": "%"
        },
        "automationbench": {
          "name": "AutomationBench",
          "value": 30.4,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 90.7,
          "unit": "%"
        },
        "aa-intelligence-index": {
          "name": "AA Intelligence Index",
          "value": 56,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1525,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 84.5,
          "unit": "%"
        },
        "gdp-pdf": {
          "name": "GDP.PDF",
          "value": 34,
          "unit": "%"
        },
        "lvbench": {
          "name": "LVBench",
          "value": 85.4,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 97,
          "unit": "%"
        },
        "mrcr-v2-1m": {
          "name": "MRCR v2 (8-needle) (1M pointwise)",
          "value": 62.5,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1588,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.7-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.7-flash"
    },
    {
      "name": "DeepSeek-V4-Pro-0813",
      "slug": "deepseek-v4-pro-0813",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2026-08-13",
      "releaseDateLabel": "Aug 13 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 87.9,
          "unit": "%"
        },
        "toolathlon-verified": {
          "name": "Toolathlon-Verified",
          "value": 74.1,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 83.3,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 42.7,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 60,
          "unit": "%"
        },
        "automationbench": {
          "name": "AutomationBench",
          "value": 31.8,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v4-pro-0813",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v4-pro-0813"
    },
    {
      "name": "Grok 4.6",
      "slug": "grok-4.6",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2026-08-12",
      "releaseDateLabel": "Aug 12 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 69.9,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 65.9,
          "unit": "%"
        },
        "frontiercode-1.1-extended": {
          "name": "FrontierCode v1.1 (Extended) (extended split)",
          "value": 61.3,
          "unit": "%"
        },
        "apex-swe": {
          "name": "APEX-SWE",
          "value": 56.4,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 92,
          "unit": "%"
        },
        "terminal-bench-3.0": {
          "name": "Terminal-Bench 3.0",
          "value": 26,
          "unit": "%"
        },
        "apex-agents": {
          "name": "APEX-Agents",
          "value": 57.5,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 15.8,
          "unit": "%"
        },
        "aa-intelligence-index": {
          "name": "AA Intelligence Index",
          "value": 61,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1753,
          "unit": "score"
        },
        "aa-briefcase": {
          "name": "AA-Briefcase",
          "value": 1577,
          "unit": "score"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.6",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.6"
    },
    {
      "name": "Muse Glimmer",
      "slug": "muse-glimmer",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2026-08-10",
      "releaseDateLabel": "Aug 10 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "30B",
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 51.2,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 76,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 51.7,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 75.5,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 22,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 83.5,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 65.9,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 953,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 78.8,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 74,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/muse-glimmer",
      "markdown": "https://aireleasetracker.com/md/model/meta/muse-glimmer"
    },
    {
      "name": "GPT-5.6-Cyber",
      "slug": "gpt-5.6-cyber",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-08-10",
      "releaseDateLabel": "Aug 10 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5.6-cyber",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.6-cyber"
    },
    {
      "name": "Muse Spark 1.2",
      "slug": "muse-spark-1.2",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2026-08-05",
      "releaseDateLabel": "Aug 5 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 59.3,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 82.9,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1498,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1544,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/muse-spark-1.2",
      "markdown": "https://aireleasetracker.com/md/model/meta/muse-spark-1.2"
    },
    {
      "name": "Qwen3.8-Max",
      "slug": "qwen3.8-max",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-08-03",
      "releaseDateLabel": "Aug 3 2026",
      "access": "proprietary",
      "type": null,
      "parameters": "2.4T",
      "contextWindow": "1M",
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 67.7,
          "unit": "%"
        },
        "paperbench": {
          "name": "PaperBench",
          "value": 93,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 86.6,
          "unit": "%"
        },
        "jobbench": {
          "name": "JobBench",
          "value": 53.4,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 86.1,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 88.4,
          "unit": "%"
        },
        "babyvision": {
          "name": "BabyVision",
          "value": 82,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1497,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1667,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.8-max",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.8-max"
    },
    {
      "name": "DeepSeek-V4-Flash-0731",
      "slug": "deepseek-v4-flash-0731",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2026-07-31",
      "releaseDateLabel": "Jul 31 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 38,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 82.7,
          "unit": "%"
        },
        "toolathlon-verified": {
          "name": "Toolathlon-Verified",
          "value": 70.3,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 76.7,
          "unit": "%"
        },
        "automationbench": {
          "name": "AutomationBench",
          "value": 25.1,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v4-flash-0731",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v4-flash-0731"
    },
    {
      "name": "Claude Opus 5",
      "slug": "claude-opus-5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-07-24",
      "releaseDateLabel": "Jul 24 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 73,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 0.2,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 1.6,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 2,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 70,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 68.8,
          "unit": "%"
        },
        "frontiercode-1.1": {
          "name": "FrontierCode v1.1 (Main) (main split)",
          "value": 53.4,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 88,
          "unit": "%"
        },
        "supabase-evals": {
          "name": "Supabase Evals (with skills)",
          "value": 95.5,
          "unit": "%"
        },
        "supabase-evals-no-skills": {
          "name": "Supabase Evals (no skills)",
          "value": 90.9,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 43.3,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 90.8,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 56.3,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 64.7,
          "unit": "%"
        },
        "arc-agi-3": {
          "name": "ARC-AGI-3",
          "value": 30.2,
          "unit": "%"
        },
        "biomysterybench-hard": {
          "name": "BioMysteryBench (hard)",
          "value": 49.4,
          "unit": "%"
        },
        "biomysterybench-human-solved": {
          "name": "BioMysteryBench (human solved)",
          "value": 90.1,
          "unit": "%"
        },
        "osworld-2.0": {
          "name": "OSWorld 2.0",
          "value": 70.6,
          "unit": "%"
        },
        "automationbench": {
          "name": "AutomationBench",
          "value": 26,
          "unit": "%"
        },
        "harvey-legal-agent-heldout": {
          "name": "Harvey's Legal Agent Benchmark (Held-out)",
          "value": 11.7,
          "unit": "%"
        },
        "healthbench-professional": {
          "name": "HealthBench Professional",
          "value": 59.8,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1861,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1495,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1673,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-5"
    },
    {
      "name": "Gemini 3.6 Flash",
      "slug": "gemini-3.6-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-07-21",
      "releaseDateLabel": "Jul 21 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 39,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 7.3,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 32.2,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 37.3,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 53.5,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 49,
          "unit": "%"
        },
        "mle-bench": {
          "name": "MLE-Bench",
          "value": 63.9,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 68,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 83,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1421,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1482,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1533,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.6-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.6-flash"
    },
    {
      "name": "Gemini 3.5 Flash-Lite",
      "slug": "gemini-3.5-flash-lite",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-07-21",
      "releaseDateLabel": "Jul 21 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 65,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 54.2,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 54,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 49,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 74,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1140,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1459,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1453,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.5-flash-lite",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.5-flash-lite"
    },
    {
      "name": "Gemini 3.5 Flash Cyber",
      "slug": "gemini-3.5-flash-cyber",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-07-21",
      "releaseDateLabel": "Jul 21 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-3.5-flash-cyber",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.5-flash-cyber"
    },
    {
      "name": "Kimi K3",
      "slug": "kimi-k3",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2026-07-16",
      "releaseDateLabel": "Jul 16 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "2.8T",
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 73,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 69,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 67.5,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 92,
          "unit": "%"
        },
        "supabase-evals": {
          "name": "Supabase Evals (with skills)",
          "value": 86.4,
          "unit": "%"
        },
        "supabase-evals-no-skills": {
          "name": "Supabase Evals (no skills)",
          "value": 90.9,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 88.3,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 84.2,
          "unit": "%"
        },
        "jobbench": {
          "name": "JobBench",
          "value": 52.9,
          "unit": "%"
        },
        "toolathlon-verified": {
          "name": "Toolathlon-Verified",
          "value": 73.2,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 91.2,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 43.5,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 56,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 93.5,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1668,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 84.8,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 81.6,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1486,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1679,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k3",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k3"
    },
    {
      "name": "Muse Spark 1.1",
      "slug": "muse-spark-1.1",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2026-07-09",
      "releaseDateLabel": "Jul 9 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 61.5,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 53.3,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 80,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 88.1,
          "unit": "%"
        },
        "jobbench": {
          "name": "JobBench",
          "value": 54.7,
          "unit": "%"
        },
        "toolathlon-verified": {
          "name": "Toolathlon-Verified",
          "value": 75.6,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 62.1,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 80.8,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 57.2,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 20,
          "unit": "%"
        },
        "taxeval-v2": {
          "name": "TaxEval v2",
          "value": 79.72,
          "unit": "%"
        },
        "medscribe": {
          "name": "MedScribe",
          "value": 88.89,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 88.4,
          "unit": "%"
        },
        "babyvision": {
          "name": "BabyVision",
          "value": 76.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1490,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1540,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/muse-spark-1.1",
      "markdown": "https://aireleasetracker.com/md/model/meta/muse-spark-1.1"
    },
    {
      "name": "Grok 4.5",
      "slug": "grok-4.5",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2026-07-08",
      "releaseDateLabel": "Jul 8 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 54,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 13.4,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 54.2,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 60.8,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 64.7,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 78,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 66.7,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 54,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 62,
          "unit": "%"
        },
        "frontiercode-1.1-extended": {
          "name": "FrontierCode v1.1 (Extended) (extended split)",
          "value": 56.6,
          "unit": "%"
        },
        "apex-swe": {
          "name": "APEX-SWE",
          "value": 53.6,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 83,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 17.8,
          "unit": "%"
        },
        "terminal-bench-3.0": {
          "name": "Terminal-Bench 3.0",
          "value": 15.7,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 83.3,
          "unit": "%"
        },
        "apex-agents": {
          "name": "APEX-Agents",
          "value": 47.1,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 12.92,
          "unit": "%"
        },
        "medscribe": {
          "name": "MedScribe",
          "value": 86.88,
          "unit": "%"
        },
        "aa-intelligence-index": {
          "name": "AA Intelligence Index",
          "value": 56,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1526,
          "unit": "score"
        },
        "aa-briefcase": {
          "name": "AA-Briefcase",
          "value": 1313,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1468,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1549,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.5",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.5"
    },
    {
      "name": "Claude Sonnet 5",
      "slug": "claude-sonnet-5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-06-30",
      "releaseDateLabel": "Jun 30 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 80,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 0.6,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 4.7,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 5.9,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 63.2,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 61.5,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 61.2,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 79,
          "unit": "%"
        },
        "supabase-evals": {
          "name": "Supabase Evals (with skills)",
          "value": 95.5,
          "unit": "%"
        },
        "supabase-evals-no-skills": {
          "name": "Supabase Evals (no skills)",
          "value": 90.9,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 14.6,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 80.4,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 84.7,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 43.2,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 57.4,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 81.2,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1618,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1463,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1543,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-sonnet-5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-sonnet-5"
    },
    {
      "name": "GPT-5.6 Sol",
      "slug": "gpt-5.6-sol",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-06-26",
      "releaseDateLabel": "Jun 26 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 47,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 3.1,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 16.3,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 20,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 67.2,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 73,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 92,
          "unit": "%"
        },
        "supabase-evals": {
          "name": "Supabase Evals (with skills)",
          "value": 95.5,
          "unit": "%"
        },
        "supabase-evals-no-skills": {
          "name": "Supabase Evals (no skills)",
          "value": 90.9,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 34.4,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 88.8,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 67,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1748,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1486,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1620,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.6-sol",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.6-sol"
    },
    {
      "name": "GPT-5.6 Terra",
      "slug": "gpt-5.6-terra",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-06-26",
      "releaseDateLabel": "Jun 26 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 53,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 5.4,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 26,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 30.4,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 64.9,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 20.8,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 84.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1467,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1526,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.6-terra",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.6-terra"
    },
    {
      "name": "GPT-5.6 Luna",
      "slug": "gpt-5.6-luna",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-06-26",
      "releaseDateLabel": "Jun 26 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 40,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 8.3,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 38.6,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 43.9,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 61.1,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 14.3,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 82.5,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1523,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.6-luna",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.6-luna"
    },
    {
      "name": "GPT-5.5-Cyber",
      "slug": "gpt-5.5-cyber",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-06-22",
      "releaseDateLabel": "Jun 22 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "cybergym": {
          "name": "CyberGym",
          "value": 85.6,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.5-cyber",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.5-cyber"
    },
    {
      "name": "GLM-5.2",
      "slug": "glm-5.2",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2026-06-16",
      "releaseDateLabel": "Jun 16 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "744B",
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 31,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 62.1,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 55,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 54.6,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 44,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 88,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 5.1,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 81,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 40.5,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 54.7,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 91.2,
          "unit": "%"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1514,
          "unit": "score"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1587,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-5.2",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-5.2"
    },
    {
      "name": "Kimi K2.7 Code",
      "slug": "kimi-k2.7-code",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2026-06-12",
      "releaseDateLabel": "Jun 12 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 49.7,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 31,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1469,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2.7-code",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2.7-code"
    },
    {
      "name": "Claude Fable 5",
      "slug": "claude-fable-5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-06-09",
      "releaseDateLabel": "Jun 9 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 54,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 0.4,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 2.3,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 2.8,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 80.3,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 95.5,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 70.5,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 72.9,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 70,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 66.1,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 92,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 33.8,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 88,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 86.9,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 64.5,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 85,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 11.25,
          "unit": "%"
        },
        "taxeval-v2": {
          "name": "TaxEval v2",
          "value": 76.94,
          "unit": "%"
        },
        "medscribe": {
          "name": "MedScribe",
          "value": 88.52,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1932,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1760,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1509,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1631,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-fable-5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-fable-5"
    },
    {
      "name": "Qwen3.7-Plus",
      "slug": "qwen3.7-plus",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-06-01",
      "releaseDateLabel": "Jun 1 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 89.6,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1458,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.7-plus",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.7-plus"
    },
    {
      "name": "Claude Opus 4.8",
      "slug": "claude-opus-4.8",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-05-28",
      "releaseDateLabel": "May 28 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 95,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 0.5,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 4.1,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 5.5,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 69.2,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 84.4,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 62.3,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 63.8,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 55.8,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 88,
          "unit": "%"
        },
        "frontier-bench-v0.1": {
          "name": "Frontier-Bench v0.1",
          "value": 21.1,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 74.6,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 74,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 84.3,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 49.8,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 57.9,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 83.4,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 53.9,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 9.58,
          "unit": "%"
        },
        "taxeval-v2": {
          "name": "TaxEval v2",
          "value": 75.63,
          "unit": "%"
        },
        "medscribe": {
          "name": "MedScribe",
          "value": 85.75,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1890,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1600,
          "unit": "score"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1482,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1568,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4.8",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4.8"
    },
    {
      "name": "Qwen3.7-Max",
      "slug": "qwen3.7-max",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-05-20",
      "releaseDateLabel": "May 20 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 71,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 60.6,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 91.6,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 69.7,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 76.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 92.4,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1475,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.7-max",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.7-max"
    },
    {
      "name": "Gemini 3.5 Flash",
      "slug": "gemini-3.5-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-05-19",
      "releaseDateLabel": "May 19 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 20,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 14.1,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 54.2,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 60.5,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 55.1,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 48.8,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 49.8,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 37,
          "unit": "%"
        },
        "mle-bench": {
          "name": "MLE-Bench",
          "value": 49.7,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 76.2,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 83.6,
          "unit": "%"
        },
        "toolathlon": {
          "name": "Toolathlon",
          "value": 56.5,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 58,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 40.2,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 72.1,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 78.4,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 57.9,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1656,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1349,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 84.2,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 83.6,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 33.6,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 77.3,
          "unit": "%"
        },
        "mrcr-v2-1m": {
          "name": "MRCR v2 (8-needle) (1M pointwise)",
          "value": 26.6,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1476,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1509,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.5-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.5-flash"
    },
    {
      "name": "Composer 2.5",
      "slug": "composer-2.5",
      "company": "cursor",
      "companyName": "Cursor",
      "releaseDate": "2026-05-18",
      "releaseDateLabel": "May 18 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 54,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 79.8,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 56.1,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 63.2,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 18,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 92,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 73,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 69.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/cursor/composer-2.5",
      "markdown": "https://aireleasetracker.com/md/model/cursor/composer-2.5"
    },
    {
      "name": "Mistral Medium 3.5",
      "slug": "mistral-medium-3.5",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2026-04-29",
      "releaseDateLabel": "Apr 29 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "128B",
      "contextWindow": "256k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 77.6,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1267,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/mistral/mistral-medium-3.5",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-medium-3.5"
    },
    {
      "name": "DeepSeek-V4-Pro",
      "slug": "deepseek-v4-pro",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2026-04-24",
      "releaseDateLabel": "Apr 24 2026",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 14,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 93.5,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 83.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 90.1,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1457,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1446,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v4-pro",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v4-pro"
    },
    {
      "name": "DeepSeek-V4-Flash",
      "slug": "deepseek-v4-flash",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2026-04-24",
      "releaseDateLabel": "Apr 24 2026",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 18,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 91.6,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 88.1,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1577,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v4-flash",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v4-flash"
    },
    {
      "name": "GPT-5.5",
      "slug": "gpt-5.5",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-04-23",
      "releaseDateLabel": "Apr 23 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1.05M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 47,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 3,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 17.4,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 20.8,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 58.6,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 77.8,
          "unit": "%"
        },
        "cursorbench-v3.2": {
          "name": "CursorBench v3.2",
          "value": 58.4,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 64.3,
          "unit": "%"
        },
        "deepswe-1.0": {
          "name": "DeepSWE 1.0",
          "value": 64.3,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 78.2,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 82.7,
          "unit": "%"
        },
        "expert-swe": {
          "name": "Expert-SWE (Internal)",
          "value": 73.1,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 75.3,
          "unit": "%"
        },
        "toolathlon": {
          "name": "Toolathlon",
          "value": 55.6,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 84.4,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 81.8,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 41.4,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 52.2,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 84.6,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 51.7,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 35.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 93.6,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 78.7,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 51.8,
          "unit": "%"
        },
        "harvey-legal-agent": {
          "name": "Harvey's Legal Agent Benchmark",
          "value": 3.75,
          "unit": "%"
        },
        "taxeval-v2": {
          "name": "TaxEval v2",
          "value": 74.98,
          "unit": "%"
        },
        "medscribe": {
          "name": "MedScribe",
          "value": 86.87,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1769,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 1494,
          "unit": "score"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 84.9,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 84.1,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 81.2,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 36.2,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 94.8,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1481,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1509,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.5",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.5"
    },
    {
      "name": "GPT-5.5-Pro",
      "slug": "gpt-5.5-pro",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-04-23",
      "releaseDateLabel": "Apr 23 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 36,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 83,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 90.1,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 57.2,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 52.4,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 39.6,
          "unit": "%"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 82.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.5-pro",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.5-pro"
    },
    {
      "name": "Kimi K2.6",
      "slug": "kimi-k2.6",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2026-04-21",
      "releaseDateLabel": "Apr 21 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 65,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 80.2,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 47.6,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 67,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 89.6,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 90.5,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1461,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1513,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2.6",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2.6"
    },
    {
      "name": "Grok 4.3 Beta",
      "slug": "grok-4.3-beta",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2026-04-17",
      "releaseDateLabel": "Apr 17 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 50,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.3-beta",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.3-beta"
    },
    {
      "name": "Claude Opus 4.7",
      "slug": "claude-opus-4.7",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-04-16",
      "releaseDateLabel": "Apr 16 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 83,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 64.3,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 87.6,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 80.5,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 64.8,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 66.1,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 69.4,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 79.1,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 79.3,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 73.1,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 46.9,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 54.7,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 75.8,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 43.8,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 22.9,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 94.2,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 78,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 51.5,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1753,
          "unit": "score"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 80.3,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 82.1,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 75.2,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 24.5,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 59.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1503,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1557,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4.7",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4.7"
    },
    {
      "name": "Qwen3.6",
      "slug": "qwen3.6",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-04-16",
      "releaseDateLabel": "Apr 16 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "35B",
      "contextWindow": "256k",
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 49.5,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 73.4,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 67.2,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 51.5,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 21.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 86,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 78,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 75.3,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 81.7,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.6",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.6"
    },
    {
      "name": "Muse Spark",
      "slug": "muse-spark",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2026-04-08",
      "releaseDateLabel": "Apr 8 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 2.9,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 14.3,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 16.5,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 55,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 77.4,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 10,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 67.3,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 82.2,
          "unit": "%"
        },
        "jobbench": {
          "name": "JobBench",
          "value": 17,
          "unit": "%"
        },
        "toolathlon-verified": {
          "name": "Toolathlon-Verified",
          "value": 49.4,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 50.4,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 42.5,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 89.5,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 53.3,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 88.9,
          "unit": "%"
        },
        "babyvision": {
          "name": "BabyVision",
          "value": 39.9,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 80.4,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1488,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/muse-spark",
      "markdown": "https://aireleasetracker.com/md/model/meta/muse-spark"
    },
    {
      "name": "GLM-5.1",
      "slug": "glm-5.1",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2026-04-07",
      "releaseDateLabel": "Apr 7 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "744B",
      "contextWindow": "200k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 22,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 63.5,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 68,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 52.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 86.2,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1468,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1518,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-5.1",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-5.1"
    },
    {
      "name": "Gemma 4",
      "slug": "gemma-4",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-04-02",
      "releaseDateLabel": "Apr 2 2026",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemma-4",
      "markdown": "https://aireleasetracker.com/md/model/google/gemma-4"
    },
    {
      "name": "Qwen3.6-Plus",
      "slug": "qwen3.6-plus",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-04-02",
      "releaseDateLabel": "Apr 2 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 72,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1458,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.6-plus",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.6-plus"
    },
    {
      "name": "Composer 2",
      "slug": "composer-2",
      "company": "cursor",
      "companyName": "Cursor",
      "releaseDate": "2026-03-19",
      "releaseDateLabel": "Mar 19 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 73.7,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 52.2,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 61.7,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/cursor/composer-2",
      "markdown": "https://aireleasetracker.com/md/model/cursor/composer-2"
    },
    {
      "name": "GPT-5.4 mini",
      "slug": "gpt-5.4-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-03-17",
      "releaseDateLabel": "Mar 17 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "400k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 32,
          "unit": "%"
        },
        "supabase-evals": {
          "name": "Supabase Evals (with skills)",
          "value": 81.8,
          "unit": "%"
        },
        "supabase-evals-no-skills": {
          "name": "Supabase Evals (no skills)",
          "value": 63.6,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 36,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1398,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.4-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.4-mini"
    },
    {
      "name": "GPT-5.4 nano",
      "slug": "gpt-5.4-nano",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-03-17",
      "releaseDateLabel": "Mar 17 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 14,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.4-nano",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.4-nano"
    },
    {
      "name": "Mistral Small 4",
      "slug": "mistral-small-4",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2026-03-16",
      "releaseDateLabel": "Mar 16 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-small-4",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-small-4"
    },
    {
      "name": "GPT-5.4",
      "slug": "gpt-5.4",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-03-05",
      "releaseDateLabel": "Mar 5 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 48,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 83,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 75.1,
          "unit": "%"
        },
        "expert-swe": {
          "name": "Expert-SWE (Internal)",
          "value": 68.5,
          "unit": "%"
        },
        "toolathlon": {
          "name": "Toolathlon",
          "value": 54.6,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 82.7,
          "unit": "%"
        },
        "cybergym": {
          "name": "CyberGym",
          "value": 79,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 47.6,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 27.1,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 92.8,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 75,
          "unit": "%"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 83,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1476,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1462,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.4",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.4"
    },
    {
      "name": "GPT-5.4-Pro",
      "slug": "gpt-5.4-pro",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-03-05",
      "releaseDateLabel": "Mar 5 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "browsecomp": {
          "name": "BrowseComp",
          "value": 89.3,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 58.7,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 83.3,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 50,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 38,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 94.4,
          "unit": "%"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 82,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.4-pro",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.4-pro"
    },
    {
      "name": "Gemini 3.1 Flash-Lite",
      "slug": "gemini-3.1-flash-lite",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-03-03",
      "releaseDateLabel": "Mar 3 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 11,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1253,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.1-flash-lite",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.1-flash-lite"
    },
    {
      "name": "Gemini 3.1 Pro",
      "slug": "gemini-3.1-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2026-02-19",
      "releaseDateLabel": "Feb 19 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 37,
          "unit": "%"
        },
        "grayswan-ipi-k1": {
          "name": "Gray Swan IPI (k = 1)",
          "value": 14.2,
          "unit": "%"
        },
        "grayswan-ipi-k10": {
          "name": "Gray Swan IPI (k = 10)",
          "value": 45.7,
          "unit": "%"
        },
        "grayswan-ipi-k15": {
          "name": "Gray Swan IPI (k = 15)",
          "value": 49.2,
          "unit": "%"
        },
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 54.2,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 80.6,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 12,
          "unit": "%"
        },
        "mle-bench": {
          "name": "MLE-Bench",
          "value": 42.6,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 70.3,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 68.5,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 78.2,
          "unit": "%"
        },
        "toolathlon": {
          "name": "Toolathlon",
          "value": 48.8,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 85.9,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 44.4,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 51.4,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 77.1,
          "unit": "%"
        },
        "frontiermath-t13": {
          "name": "FrontierMath (Tier 1–3)",
          "value": 36.9,
          "unit": "%"
        },
        "frontiermath-t4": {
          "name": "FrontierMath (Tier 4)",
          "value": 16.7,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 94.3,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 76.2,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 43,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1314,
          "unit": "score"
        },
        "gdpval-aa-v2": {
          "name": "GDPval-AA v2",
          "value": 965,
          "unit": "score"
        },
        "gdpval-winrate": {
          "name": "GDPval (win/tie rate)",
          "value": 67.3,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 83.3,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 80.5,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 26.5,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 84.9,
          "unit": "%"
        },
        "mrcr-v2-1m": {
          "name": "MRCR v2 (8-needle) (1M pointwise)",
          "value": 26.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1485,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1445,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.1-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.1-pro"
    },
    {
      "name": "Grok 4.20 Beta",
      "slug": "grok-4.20-beta",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2026-02-17",
      "releaseDateLabel": "Feb 17 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 56,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 53.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1475,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.20-beta",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.20-beta"
    },
    {
      "name": "Claude Sonnet 4.6",
      "slug": "claude-sonnet-4.6",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-02-17",
      "releaseDateLabel": "Feb 17 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 91,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 79.6,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 49,
          "unit": "%"
        },
        "deepswe-1.1": {
          "name": "DeepSWE 1.1",
          "value": 30,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 58,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 69.5,
          "unit": "%"
        },
        "bu-bench": {
          "name": "BU Bench",
          "value": 62,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 33.2,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 58.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 89.9,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 72.5,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 51,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1676,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 72.4,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 74.5,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 6.7,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 84.9,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1472,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1521,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-sonnet-4.6",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-sonnet-4.6"
    },
    {
      "name": "Qwen3.5",
      "slug": "qwen3.5",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-02-16",
      "releaseDateLabel": "Feb 16 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "397B",
      "contextWindow": "1M",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 76.4,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 69.3,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 52.5,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 69,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 28.7,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 88.4,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 62.2,
          "unit": "%"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 80.8,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 79,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 85,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3.5",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.5"
    },
    {
      "name": "Qwen3.5-Plus",
      "slug": "qwen3.5-plus",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-02-16",
      "releaseDateLabel": "Feb 16 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen3.5-plus",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3.5-plus"
    },
    {
      "name": "GPT-5.3-Codex-Spark",
      "slug": "gpt-5.3-codex-spark",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-02-12",
      "releaseDateLabel": "Feb 12 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5.3-codex-spark",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.3-codex-spark"
    },
    {
      "name": "GLM-5",
      "slug": "glm-5",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2026-02-12",
      "releaseDateLabel": "Feb 12 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "744B",
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 28,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 77.8,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 73.3,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 56.2,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 75.9,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 50.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 86,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1435,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-5",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-5"
    },
    {
      "name": "Composer 1.5",
      "slug": "composer-1.5",
      "company": "cursor",
      "companyName": "Cursor",
      "releaseDate": "2026-02-09",
      "releaseDateLabel": "Feb 9 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 67,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/cursor/composer-1.5",
      "markdown": "https://aireleasetracker.com/md/model/cursor/composer-1.5"
    },
    {
      "name": "Claude Opus 4.6",
      "slug": "claude-opus-4.6",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2026-02-05",
      "releaseDateLabel": "Feb 5 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 87,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 80.8,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 75,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 83.7,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 53,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 91.3,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1504,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1543,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4.6",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4.6"
    },
    {
      "name": "GPT-5.3-Codex",
      "slug": "gpt-5.3-codex",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2026-02-05",
      "releaseDateLabel": "Feb 5 2026",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 24,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 83,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1406,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.3-codex",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.3-codex"
    },
    {
      "name": "Qwen3-Coder-Next",
      "slug": "qwen3-coder-next",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2026-02-03",
      "releaseDateLabel": "Feb 3 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "80B",
      "contextWindow": "256k",
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 44.3,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 70.6,
          "unit": "%"
        },
        "swe-bench-multilingual": {
          "name": "SWE-Bench Multilingual",
          "value": 62.8,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 36.2,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3-coder-next",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3-coder-next"
    },
    {
      "name": "Kimi K2.5",
      "slug": "kimi-k2.5",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2026-01-27",
      "releaseDateLabel": "Jan 27 2026",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 52,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 76.8,
          "unit": "%"
        },
        "cursorbench-v3.1": {
          "name": "CursorBench v3.1",
          "value": 31.9,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 21,
          "unit": "%"
        },
        "livecodebench": {
          "name": "LiveCodeBench",
          "value": 85,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 87.6,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1433,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2.5",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2.5"
    },
    {
      "name": "GLM-4.7",
      "slug": "glm-4.7",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2025-12-22",
      "releaseDateLabel": "Dec 22 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 73.8,
          "unit": "%"
        },
        "terminal-bench-2.0": {
          "name": "Terminal-Bench 2.0",
          "value": 41,
          "unit": "%"
        },
        "browsecomp": {
          "name": "BrowseComp",
          "value": 52,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 24.8,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 42.8,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 85.7,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1433,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-4.7",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4.7"
    },
    {
      "name": "Gemini 3.0 Flash",
      "slug": "gemini-3.0-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-12-17",
      "releaseDateLabel": "Dec 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-pro": {
          "name": "SWE-Bench Pro",
          "value": 49.6,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 78,
          "unit": "%"
        },
        "terminal-bench-2.1": {
          "name": "Terminal-Bench 2.1",
          "value": 58,
          "unit": "%"
        },
        "mcp-atlas": {
          "name": "MCP Atlas",
          "value": 62,
          "unit": "%"
        },
        "toolathlon": {
          "name": "Toolathlon",
          "value": 49.4,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 33.7,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 33.6,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 90.4,
          "unit": "%"
        },
        "osworld-verified": {
          "name": "OSWorld-Verified",
          "value": 65.1,
          "unit": "%"
        },
        "finance-agent-v2": {
          "name": "Finance Agent v2",
          "value": 42.6,
          "unit": "%"
        },
        "gdpval-aa": {
          "name": "GDPval-AA",
          "value": 1204,
          "unit": "score"
        },
        "charxiv-reasoning": {
          "name": "CharXiv Reasoning",
          "value": 80.3,
          "unit": "%"
        },
        "mmmu-pro": {
          "name": "MMMU-Pro",
          "value": 81.2,
          "unit": "%"
        },
        "blueprint-bench-2": {
          "name": "Blueprint-Bench 2",
          "value": 0,
          "unit": "%"
        },
        "mrcr-v2-128k": {
          "name": "MRCR v2 (8-needle) (128k average)",
          "value": 67.2,
          "unit": "%"
        },
        "mrcr-v2-1m": {
          "name": "MRCR v2 (8-needle) (1M pointwise)",
          "value": 22.1,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1473,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1437,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.0-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.0-flash"
    },
    {
      "name": "GPT-5.2",
      "slug": "gpt-5.2",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-12-11",
      "releaseDateLabel": "Dec 11 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 38,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 80,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 52.9,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 92.4,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1476,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1419,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.2",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.2"
    },
    {
      "name": "Devstral 2 (123B)",
      "slug": "devstral-2-(123b)",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-09",
      "releaseDateLabel": "Dec 9 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/devstral-2-(123b)",
      "markdown": "https://aireleasetracker.com/md/model/mistral/devstral-2-(123b)"
    },
    {
      "name": "Devstral Small 2 (24B)",
      "slug": "devstral-small-2-(24b)",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-09",
      "releaseDateLabel": "Dec 9 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/devstral-small-2-(24b)",
      "markdown": "https://aireleasetracker.com/md/model/mistral/devstral-small-2-(24b)"
    },
    {
      "name": "Ministral 3 3B-25-12",
      "slug": "ministral-3-3b-25-12",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-02",
      "releaseDateLabel": "Dec 2 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/ministral-3-3b-25-12",
      "markdown": "https://aireleasetracker.com/md/model/mistral/ministral-3-3b-25-12"
    },
    {
      "name": "Ministral 3 8B-25-12",
      "slug": "ministral-3-8b-25-12",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-02",
      "releaseDateLabel": "Dec 2 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/ministral-3-8b-25-12",
      "markdown": "https://aireleasetracker.com/md/model/mistral/ministral-3-8b-25-12"
    },
    {
      "name": "Ministral 3 14B-25-12",
      "slug": "ministral-3-14b-25-12",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-02",
      "releaseDateLabel": "Dec 2 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/ministral-3-14b-25-12",
      "markdown": "https://aireleasetracker.com/md/model/mistral/ministral-3-14b-25-12"
    },
    {
      "name": "Mistral Large 3",
      "slug": "mistral-large-3",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-12-02",
      "releaseDateLabel": "Dec 2 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1230,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/mistral/mistral-large-3",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-large-3"
    },
    {
      "name": "DeepSeek-V3.2",
      "slug": "deepseek-v3.2",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-12-01",
      "releaseDateLabel": "Dec 1 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 13,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1361,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3.2",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3.2"
    },
    {
      "name": "Claude Opus 4.5",
      "slug": "claude-opus-4.5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-11-24",
      "releaseDateLabel": "Nov 24 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 90,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 80.9,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 87,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4.5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4.5"
    },
    {
      "name": "Grok 4.1 Fast",
      "slug": "grok-4.1-fast",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-11-19",
      "releaseDateLabel": "Nov 19 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 19,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.1-fast",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.1-fast"
    },
    {
      "name": "GPT-5.1-Codex-Max",
      "slug": "gpt-5.1-codex-max",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-11-19",
      "releaseDateLabel": "Nov 19 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5.1-codex-max",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.1-codex-max"
    },
    {
      "name": "Gemini 3.0 Pro",
      "slug": "gemini-3.0-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-11-18",
      "releaseDateLabel": "Nov 18 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 76.2,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 67,
          "unit": "%"
        },
        "arc-agi-2": {
          "name": "ARC-AGI-2",
          "value": 31.1,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 91.9,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 81,
          "unit": "%"
        },
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1486,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1439,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-3.0-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-3.0-pro"
    },
    {
      "name": "Grok 4.1",
      "slug": "grok-4.1",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-11-17",
      "releaseDateLabel": "Nov 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "arena-text-elo": {
          "name": "Arena Elo (Text)",
          "value": 1466,
          "unit": "elo"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1209,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-4.1",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4.1"
    },
    {
      "name": "GPT-5.1",
      "slug": "gpt-5.1",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-11-12",
      "releaseDateLabel": "Nov 12 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 25,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 76.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 88.1,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 76,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1391,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5.1",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5.1"
    },
    {
      "name": "GPT-5-Codex-Mini",
      "slug": "gpt-5-codex-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-11-07",
      "releaseDateLabel": "Nov 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5-codex-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5-codex-mini"
    },
    {
      "name": "Kimi K2 Thinking",
      "slug": "kimi-k2-thinking",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-11-06",
      "releaseDateLabel": "Nov 6 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 71.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2-thinking",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2-thinking"
    },
    {
      "name": "Kimi Linear",
      "slug": "kimi-linear",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-10-30",
      "releaseDateLabel": "Oct 30 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "48B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/moonshot/kimi-linear",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-linear"
    },
    {
      "name": "Composer",
      "slug": "composer",
      "company": "cursor",
      "companyName": "Cursor",
      "releaseDate": "2025-10-29",
      "releaseDateLabel": "Oct 29 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/cursor/composer",
      "markdown": "https://aireleasetracker.com/md/model/cursor/composer"
    },
    {
      "name": "Claude Haiku 4.5",
      "slug": "claude-haiku-4.5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-10-15",
      "releaseDateLabel": "Oct 15 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 77,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 73.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 73,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-haiku-4.5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-haiku-4.5"
    },
    {
      "name": "GLM-4.6",
      "slug": "glm-4.6",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2025-09-30",
      "releaseDateLabel": "Sep 30 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "355B",
      "contextWindow": "200k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 68,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1339,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-4.6",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4.6"
    },
    {
      "name": "Claude Sonnet 4.5",
      "slug": "claude-sonnet-4.5",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-09-29",
      "releaseDateLabel": "Sep 29 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 79,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 77.2,
          "unit": "%"
        },
        "nextjs-evals": {
          "name": "Next.js Evals",
          "value": 50,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 83.4,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 68,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-sonnet-4.5",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-sonnet-4.5"
    },
    {
      "name": "DeepSeek V3.2 Exp",
      "slug": "deepseek-v3.2-exp",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-09-29",
      "releaseDateLabel": "Sep 29 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1272,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3.2-exp",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3.2-exp"
    },
    {
      "name": "Qwen3-Max",
      "slug": "qwen3-max",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-09-24",
      "releaseDateLabel": "Sep 24 2025",
      "access": "proprietary",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen3-max",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3-max"
    },
    {
      "name": "DeepSeek V3.1 Terminus",
      "slug": "deepseek-v3.1-terminus",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-09-22",
      "releaseDateLabel": "Sep 22 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3.1-terminus",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3.1-terminus"
    },
    {
      "name": "Grok 4 Fast",
      "slug": "grok-4-fast",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-09-19",
      "releaseDateLabel": "Sep 19 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-4-fast",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4-fast"
    },
    {
      "name": "GPT-5-Codex",
      "slug": "gpt-5-codex",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-09-15",
      "releaseDateLabel": "Sep 15 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 39,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5-codex",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5-codex"
    },
    {
      "name": "Qwen3-Next",
      "slug": "qwen3-next",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-09-11",
      "releaseDateLabel": "Sep 11 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "80B",
      "contextWindow": "256k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen3-next",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3-next"
    },
    {
      "name": "Kimi K2 (0905)",
      "slug": "kimi-k2-(0905)",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-09-05",
      "releaseDateLabel": "Sep 5 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "256k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 65.8,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 75.1,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2-(0905)",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2-(0905)"
    },
    {
      "name": "Grok Code Fast 1",
      "slug": "grok-code-fast-1",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-08-28",
      "releaseDateLabel": "Aug 28 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1163,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/xai/grok-code-fast-1",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-code-fast-1"
    },
    {
      "name": "Grok 2.5",
      "slug": "grok-2.5",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-08-23",
      "releaseDateLabel": "Aug 23 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-2.5",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-2.5"
    },
    {
      "name": "DeepSeek V3.1",
      "slug": "deepseek-v3.1",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-08-21",
      "releaseDateLabel": "Aug 21 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3.1",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3.1"
    },
    {
      "name": "GPT-5",
      "slug": "gpt-5",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-07",
      "releaseDateLabel": "Aug 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 21,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1419,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-5",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5"
    },
    {
      "name": "GPT-5 mini",
      "slug": "gpt-5-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-07",
      "releaseDateLabel": "Aug 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5-mini"
    },
    {
      "name": "GPT-5 nano",
      "slug": "gpt-5-nano",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-07",
      "releaseDateLabel": "Aug 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5-nano",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5-nano"
    },
    {
      "name": "GPT-5 Pro",
      "slug": "gpt-5-pro",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-07",
      "releaseDateLabel": "Aug 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-5-pro",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-5-pro"
    },
    {
      "name": "Claude Opus 4.1",
      "slug": "claude-opus-4.1",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-08-05",
      "releaseDateLabel": "Aug 5 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 43,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 74.5,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 80.9,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4.1",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4.1"
    },
    {
      "name": "gpt-oss-120b",
      "slug": "gpt-oss-120b",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-05",
      "releaseDateLabel": "Aug 5 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "117B",
      "contextWindow": "128k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 11,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 62.4,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 14.9,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 19,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 80.1,
          "unit": "%"
        },
        "mmlu": {
          "name": "MMLU",
          "value": 90,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-oss-120b",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-oss-120b"
    },
    {
      "name": "gpt-oss-20b",
      "slug": "gpt-oss-20b",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-08-05",
      "releaseDateLabel": "Aug 5 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "21B",
      "contextWindow": "128k",
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 60.7,
          "unit": "%"
        },
        "hle-no-tools": {
          "name": "Humanity's Last Exam (no tools)",
          "value": 10.9,
          "unit": "%"
        },
        "hle-with-tools": {
          "name": "Humanity's Last Exam (with tools)",
          "value": 17.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 71.5,
          "unit": "%"
        },
        "mmlu": {
          "name": "MMLU",
          "value": 85.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-oss-20b",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-oss-20b"
    },
    {
      "name": "GLM-4.5",
      "slug": "glm-4.5",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2025-07-28",
      "releaseDateLabel": "Jul 28 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "355B",
      "contextWindow": "128k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 8,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 64.2,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 79.1,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-4.5",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4.5"
    },
    {
      "name": "GLM-4.5-Air",
      "slug": "glm-4.5-air",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2025-07-28",
      "releaseDateLabel": "Jul 28 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "106B",
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 59.8,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/zai/glm-4.5-air",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4.5-air"
    },
    {
      "name": "Qwen3-Coder",
      "slug": "qwen3-coder",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-07-22",
      "releaseDateLabel": "Jul 22 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "480B",
      "contextWindow": "256k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 20,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/qwen/qwen3-coder",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3-coder"
    },
    {
      "name": "Kimi K2",
      "slug": "kimi-k2",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-07-11",
      "releaseDateLabel": "Jul 11 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "1T",
      "contextWindow": "128k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 10,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 65.8,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 75.1,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k2",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k2"
    },
    {
      "name": "Grok 4",
      "slug": "grok-4",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-07-09",
      "releaseDateLabel": "Jul 9 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-4",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4"
    },
    {
      "name": "Grok 4 Heavy",
      "slug": "grok-4-heavy",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-07-09",
      "releaseDateLabel": "Jul 9 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-4-heavy",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-4-heavy"
    },
    {
      "name": "Mistral Small 3.2",
      "slug": "mistral-small-3.2",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-06-20",
      "releaseDateLabel": "Jun 20 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "24B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-small-3.2",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-small-3.2"
    },
    {
      "name": "Gemini 2.5 Flash-Lite",
      "slug": "gemini-2.5-flash-lite",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-06-17",
      "releaseDateLabel": "Jun 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-2.5-flash-lite",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.5-flash-lite"
    },
    {
      "name": "Kimi-Dev",
      "slug": "kimi-dev",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-06-17",
      "releaseDateLabel": "Jun 17 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "72B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/moonshot/kimi-dev",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-dev"
    },
    {
      "name": "o3-pro",
      "slug": "o3-pro",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-06-10",
      "releaseDateLabel": "Jun 10 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/o3-pro",
      "markdown": "https://aireleasetracker.com/md/model/openai/o3-pro"
    },
    {
      "name": "Magistral Small",
      "slug": "magistral-small",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-06-10",
      "releaseDateLabel": "Jun 10 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/magistral-small",
      "markdown": "https://aireleasetracker.com/md/model/mistral/magistral-small"
    },
    {
      "name": "Magistral Medium",
      "slug": "magistral-medium",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-06-10",
      "releaseDateLabel": "Jun 10 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/magistral-medium",
      "markdown": "https://aireleasetracker.com/md/model/mistral/magistral-medium"
    },
    {
      "name": "DeepSeek R1-0528",
      "slug": "deepseek-r1-0528",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-05-28",
      "releaseDateLabel": "May 28 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 8,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-r1-0528",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-r1-0528"
    },
    {
      "name": "Claude Sonnet 4",
      "slug": "claude-sonnet-4",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-05-22",
      "releaseDateLabel": "May 22 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 30,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 72.7,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 75.4,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-sonnet-4",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-sonnet-4"
    },
    {
      "name": "Claude Opus 4",
      "slug": "claude-opus-4",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-05-22",
      "releaseDateLabel": "May 22 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 34,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 72.5,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 79.6,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-opus-4",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-opus-4"
    },
    {
      "name": "Devstral Small",
      "slug": "devstral-small",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-05-21",
      "releaseDateLabel": "May 21 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "24B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/devstral-small",
      "markdown": "https://aireleasetracker.com/md/model/mistral/devstral-small"
    },
    {
      "name": "Mistral Medium 3",
      "slug": "mistral-medium-3",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-05-07",
      "releaseDateLabel": "May 7 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-medium-3",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-medium-3"
    },
    {
      "name": "Qwen3",
      "slug": "qwen3",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-04-29",
      "releaseDateLabel": "Apr 29 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "235B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen3",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen3"
    },
    {
      "name": "Gemini 2.5 Flash",
      "slug": "gemini-2.5-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-04-17",
      "releaseDateLabel": "Apr 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 19,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-2.5-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.5-flash"
    },
    {
      "name": "o3",
      "slug": "o3",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-16",
      "releaseDateLabel": "Apr 16 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 26,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/o3",
      "markdown": "https://aireleasetracker.com/md/model/openai/o3"
    },
    {
      "name": "o4-mini",
      "slug": "o4-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-16",
      "releaseDateLabel": "Apr 16 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 8,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/o4-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/o4-mini"
    },
    {
      "name": "o4-mini-high",
      "slug": "o4-mini-high",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-16",
      "releaseDateLabel": "Apr 16 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/o4-mini-high",
      "markdown": "https://aireleasetracker.com/md/model/openai/o4-mini-high"
    },
    {
      "name": "GPT-4.1",
      "slug": "gpt-4.1",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-14",
      "releaseDateLabel": "Apr 14 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 14,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-4.1",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4.1"
    },
    {
      "name": "GPT-4.1 mini",
      "slug": "gpt-4.1-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-14",
      "releaseDateLabel": "Apr 14 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-4.1-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4.1-mini"
    },
    {
      "name": "GPT-4.1 nano",
      "slug": "gpt-4.1-nano",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-04-14",
      "releaseDateLabel": "Apr 14 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "1M",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-4.1-nano",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4.1-nano"
    },
    {
      "name": "Kimi-VL",
      "slug": "kimi-vl",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-04-10",
      "releaseDateLabel": "Apr 10 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "16B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/moonshot/kimi-vl",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-vl"
    },
    {
      "name": "LLaMA 4 Scout",
      "slug": "llama-4-scout",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2025-04-05",
      "releaseDateLabel": "Apr 5 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 19,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/llama-4-scout",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-4-scout"
    },
    {
      "name": "LLaMA 4 Maverick",
      "slug": "llama-4-maverick",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2025-04-05",
      "releaseDateLabel": "Apr 5 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 28,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/llama-4-maverick",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-4-maverick"
    },
    {
      "name": "Gemini 2.5 Pro",
      "slug": "gemini-2.5-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-03-25",
      "releaseDateLabel": "Mar 25 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 20,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 59.6,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 86.4,
          "unit": "%"
        },
        "mmmu": {
          "name": "MMMU",
          "value": 68,
          "unit": "%"
        },
        "arena-code-elo": {
          "name": "Arena Elo (Code)",
          "value": 1224,
          "unit": "elo"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-2.5-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.5-pro"
    },
    {
      "name": "DeepSeek V3-0324",
      "slug": "deepseek-v3-0324",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-03-24",
      "releaseDateLabel": "Mar 24 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3-0324",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3-0324"
    },
    {
      "name": "o1-pro",
      "slug": "o1-pro",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-03-19",
      "releaseDateLabel": "Mar 19 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/o1-pro",
      "markdown": "https://aireleasetracker.com/md/model/openai/o1-pro"
    },
    {
      "name": "Mistral Small 3.1",
      "slug": "mistral-small-3.1",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-03-17",
      "releaseDateLabel": "Mar 17 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-small-3.1",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-small-3.1"
    },
    {
      "name": "Gemma 3",
      "slug": "gemma-3",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-03-12",
      "releaseDateLabel": "Mar 12 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemma-3",
      "markdown": "https://aireleasetracker.com/md/model/google/gemma-3"
    },
    {
      "name": "QwQ-32B",
      "slug": "qwq-32b",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-03-06",
      "releaseDateLabel": "Mar 6 2025",
      "access": "open-weight",
      "type": null,
      "parameters": "32B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwq-32b",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwq-32b"
    },
    {
      "name": "GPT-4.5",
      "slug": "gpt-4.5",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-02-27",
      "releaseDateLabel": "Feb 27 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-4.5",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4.5"
    },
    {
      "name": "Claude 3.7 Sonnet",
      "slug": "claude-3.7-sonnet",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2025-02-24",
      "releaseDateLabel": "Feb 24 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 49,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 62.3,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 68,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3.7-sonnet",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3.7-sonnet"
    },
    {
      "name": "Grok 3",
      "slug": "grok-3",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-02-17",
      "releaseDateLabel": "Feb 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-3",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-3"
    },
    {
      "name": "Grok 3 mini",
      "slug": "grok-3-mini",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2025-02-17",
      "releaseDateLabel": "Feb 17 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok-3-mini",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok-3-mini"
    },
    {
      "name": "Gemini 2.0 Flash-Lite",
      "slug": "gemini-2.0-flash-lite",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-02-05",
      "releaseDateLabel": "Feb 5 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-2.0-flash-lite",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.0-flash-lite"
    },
    {
      "name": "Gemini 2.0 Pro",
      "slug": "gemini-2.0-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-02-05",
      "releaseDateLabel": "Feb 5 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-2.0-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.0-pro"
    },
    {
      "name": "o3-mini",
      "slug": "o3-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-01-31",
      "releaseDateLabel": "Jan 31 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/o3-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/o3-mini"
    },
    {
      "name": "o3-mini-high",
      "slug": "o3-mini-high",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2025-01-31",
      "releaseDateLabel": "Jan 31 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/o3-mini-high",
      "markdown": "https://aireleasetracker.com/md/model/openai/o3-mini-high"
    },
    {
      "name": "Gemini 2.0 Flash",
      "slug": "gemini-2.0-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2025-01-30",
      "releaseDateLabel": "Jan 30 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 15,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/google/gemini-2.0-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-2.0-flash"
    },
    {
      "name": "Mistral Small 3",
      "slug": "mistral-small-3",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-01-30",
      "releaseDateLabel": "Jan 30 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-small-3",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-small-3"
    },
    {
      "name": "Qwen2.5-Max",
      "slug": "qwen2.5-max",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2025-01-28",
      "releaseDateLabel": "Jan 28 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "32k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen2.5-max",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen2.5-max"
    },
    {
      "name": "DeepSeek Chat (R1-based)",
      "slug": "deepseek-chat-(r1-based)",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2025-01-20",
      "releaseDateLabel": "Jan 20 2025",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-chat-(r1-based)",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-chat-(r1-based)"
    },
    {
      "name": "Kimi K1.5",
      "slug": "kimi-k1.5",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2025-01-20",
      "releaseDateLabel": "Jan 20 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/moonshot/kimi-k1.5",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-k1.5"
    },
    {
      "name": "Codestral 25.01",
      "slug": "codestral-25.01",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2025-01-13",
      "releaseDateLabel": "Jan 13 2025",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "256k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/codestral-25.01",
      "markdown": "https://aireleasetracker.com/md/model/mistral/codestral-25.01"
    },
    {
      "name": "DeepSeek V3 Chat",
      "slug": "deepseek-v3-chat",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-12-26",
      "releaseDateLabel": "Dec 26 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3-chat",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3-chat"
    },
    {
      "name": "DeepSeek V3 Base",
      "slug": "deepseek-v3-base",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-12-25",
      "releaseDateLabel": "Dec 25 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v3-base",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v3-base"
    },
    {
      "name": "DeepSeek V2.5 (Revised)",
      "slug": "deepseek-v2.5-(revised)",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-12-10",
      "releaseDateLabel": "Dec 10 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v2.5-(revised)",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v2.5-(revised)"
    },
    {
      "name": "LLaMA 3.3",
      "slug": "llama-3.3",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2024-12-06",
      "releaseDateLabel": "Dec 6 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "70B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/llama-3.3",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-3.3"
    },
    {
      "name": "o1",
      "slug": "o1",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2024-12-05",
      "releaseDateLabel": "Dec 5 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "200k",
      "benchmarks": {
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 75.7,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/o1",
      "markdown": "https://aireleasetracker.com/md/model/openai/o1"
    },
    {
      "name": "QwQ-32B-Preview",
      "slug": "qwq-32b-preview",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2024-11-28",
      "releaseDateLabel": "Nov 28 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "32B",
      "contextWindow": "32k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwq-32b-preview",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwq-32b-preview"
    },
    {
      "name": "DeepSeek-R1-Lite Preview",
      "slug": "deepseek-r1-lite-preview",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-11-20",
      "releaseDateLabel": "Nov 20 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-r1-lite-preview",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-r1-lite-preview"
    },
    {
      "name": "Mistral Large 2 24.11",
      "slug": "mistral-large-2-24.11",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-11-18",
      "releaseDateLabel": "Nov 18 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-large-2-24.11",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-large-2-24.11"
    },
    {
      "name": "Pixtral Large 24.11",
      "slug": "pixtral-large-24.11",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-11-18",
      "releaseDateLabel": "Nov 18 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/pixtral-large-24.11",
      "markdown": "https://aireleasetracker.com/md/model/mistral/pixtral-large-24.11"
    },
    {
      "name": "Qwen2.5-Coder",
      "slug": "qwen2.5-coder",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2024-11-12",
      "releaseDateLabel": "Nov 12 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "32B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen2.5-coder",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen2.5-coder"
    },
    {
      "name": "Claude 3.5 Haiku",
      "slug": "claude-3.5-haiku",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-10-22",
      "releaseDateLabel": "Oct 22 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 50,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 40.6,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 41.6,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3.5-haiku",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3.5-haiku"
    },
    {
      "name": "Claude 3.5 Sonnet (upgraded)",
      "slug": "claude-3.5-sonnet-(upgraded)",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-10-22",
      "releaseDateLabel": "Oct 22 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 49,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3.5-sonnet-(upgraded)",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3.5-sonnet-(upgraded)"
    },
    {
      "name": "Ministral 8B",
      "slug": "ministral-8b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-10-16",
      "releaseDateLabel": "Oct 16 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/ministral-8b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/ministral-8b"
    },
    {
      "name": "Ministral 3B",
      "slug": "ministral-3b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-10-16",
      "releaseDateLabel": "Oct 16 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/ministral-3b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/ministral-3b"
    },
    {
      "name": "Gemini 1.5 Flash-8B",
      "slug": "gemini-1.5-flash-8b",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-10-03",
      "releaseDateLabel": "Oct 3 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.5-flash-8b",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.5-flash-8b"
    },
    {
      "name": "LLaMA 3.2",
      "slug": "llama-3.2",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2024-09-25",
      "releaseDateLabel": "Sep 25 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "1B/3B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/llama-3.2",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-3.2"
    },
    {
      "name": "Gemini 1.5 Pro-002",
      "slug": "gemini-1.5-pro-002",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-09-24",
      "releaseDateLabel": "Sep 24 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.5-pro-002",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.5-pro-002"
    },
    {
      "name": "Gemini 1.5 Flash-002",
      "slug": "gemini-1.5-flash-002",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-09-24",
      "releaseDateLabel": "Sep 24 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.5-flash-002",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.5-flash-002"
    },
    {
      "name": "Qwen2.5",
      "slug": "qwen2.5",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2024-09-19",
      "releaseDateLabel": "Sep 19 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "72B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen2.5",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen2.5"
    },
    {
      "name": "o1-preview",
      "slug": "o1-preview",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2024-09-12",
      "releaseDateLabel": "Sep 12 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 73.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/o1-preview",
      "markdown": "https://aireleasetracker.com/md/model/openai/o1-preview"
    },
    {
      "name": "o1-mini",
      "slug": "o1-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2024-09-12",
      "releaseDateLabel": "Sep 12 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 60,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/o1-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/o1-mini"
    },
    {
      "name": "Pixtral 24.09",
      "slug": "pixtral-24.09",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-09-11",
      "releaseDateLabel": "Sep 11 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/pixtral-24.09",
      "markdown": "https://aireleasetracker.com/md/model/mistral/pixtral-24.09"
    },
    {
      "name": "DeepSeek V2.5",
      "slug": "deepseek-v2.5",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-09-05",
      "releaseDateLabel": "Sep 5 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v2.5",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v2.5"
    },
    {
      "name": "GLM-4-Plus",
      "slug": "glm-4-plus",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2024-08-29",
      "releaseDateLabel": "Aug 29 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/zai/glm-4-plus",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4-plus"
    },
    {
      "name": "Grok‑2",
      "slug": "grok‑2",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2024-08-14",
      "releaseDateLabel": "Aug 14 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok‑2",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok‑2"
    },
    {
      "name": "Mistral Large 2",
      "slug": "mistral-large-2",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-07-24",
      "releaseDateLabel": "Jul 24 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-large-2",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-large-2"
    },
    {
      "name": "LLaMA 3.1",
      "slug": "llama-3.1",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2024-07-23",
      "releaseDateLabel": "Jul 23 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 14,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/meta/llama-3.1",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-3.1"
    },
    {
      "name": "GPT-4o mini",
      "slug": "gpt-4o-mini",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2024-07-18",
      "releaseDateLabel": "Jul 18 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 2,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 40.2,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-4o-mini",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4o-mini"
    },
    {
      "name": "Mistral NeMo",
      "slug": "mistral-nemo",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-07-18",
      "releaseDateLabel": "Jul 18 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "12B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-nemo",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-nemo"
    },
    {
      "name": "Codestral Mamba 7B",
      "slug": "codestral-mamba-7b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-07-16",
      "releaseDateLabel": "Jul 16 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/codestral-mamba-7b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/codestral-mamba-7b"
    },
    {
      "name": "Mathstral 7B",
      "slug": "mathstral-7b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-07-16",
      "releaseDateLabel": "Jul 16 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mathstral-7b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mathstral-7b"
    },
    {
      "name": "Gemma 2",
      "slug": "gemma-2",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-06-27",
      "releaseDateLabel": "Jun 27 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemma-2",
      "markdown": "https://aireleasetracker.com/md/model/google/gemma-2"
    },
    {
      "name": "Claude 3.5 Sonnet",
      "slug": "claude-3.5-sonnet",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-06-20",
      "releaseDateLabel": "Jun 20 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 45,
          "unit": "%"
        },
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 33.4,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 59.4,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3.5-sonnet",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3.5-sonnet"
    },
    {
      "name": "DeepSeek Coder V2",
      "slug": "deepseek-coder-v2",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-06-17",
      "releaseDateLabel": "Jun 17 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-coder-v2",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-coder-v2"
    },
    {
      "name": "Qwen2",
      "slug": "qwen2",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2024-06-06",
      "releaseDateLabel": "Jun 6 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "72B",
      "contextWindow": "128k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen2",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen2"
    },
    {
      "name": "GLM-4-9B",
      "slug": "glm-4-9b",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2024-06-05",
      "releaseDateLabel": "Jun 5 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "9B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/zai/glm-4-9b",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4-9b"
    },
    {
      "name": "Codestral 22B",
      "slug": "codestral-22b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-05-29",
      "releaseDateLabel": "May 29 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/codestral-22b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/codestral-22b"
    },
    {
      "name": "Gemini 1.5 Flash",
      "slug": "gemini-1.5-flash",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-05-14",
      "releaseDateLabel": "May 14 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.5-flash",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.5-flash"
    },
    {
      "name": "GPT-4o",
      "slug": "gpt-4o",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2024-05-13",
      "releaseDateLabel": "May 13 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 12,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 49.9,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-4o",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4o"
    },
    {
      "name": "DeepSeek V2",
      "slug": "deepseek-v2",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-05-06",
      "releaseDateLabel": "May 6 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-v2",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-v2"
    },
    {
      "name": "LLaMA 3 (8B/70B)",
      "slug": "llama-3-(8b/70b)",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2024-04-18",
      "releaseDateLabel": "Apr 18 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/llama-3-(8b/70b)",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-3-(8b/70b)"
    },
    {
      "name": "Mixtral 8×22B",
      "slug": "mixtral-8×22b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-04-10",
      "releaseDateLabel": "Apr 10 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mixtral-8×22b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mixtral-8×22b"
    },
    {
      "name": "Grok‑1.5",
      "slug": "grok‑1.5",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2024-03-29",
      "releaseDateLabel": "Mar 29 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok‑1.5",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok‑1.5"
    },
    {
      "name": "Claude 3 Haiku",
      "slug": "claude-3-haiku",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-03-04",
      "releaseDateLabel": "Mar 4 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "bullshitbench-v2": {
          "name": "BullshitBench v2",
          "value": 10,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 33.3,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3-haiku",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3-haiku"
    },
    {
      "name": "Claude 3 Sonnet",
      "slug": "claude-3-sonnet",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-03-04",
      "releaseDateLabel": "Mar 4 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 40.4,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3-sonnet",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3-sonnet"
    },
    {
      "name": "Claude 3 Opus",
      "slug": "claude-3-opus",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2024-03-04",
      "releaseDateLabel": "Mar 4 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {
        "swe-bench-verified": {
          "name": "SWE-Bench Verified",
          "value": 33,
          "unit": "%"
        },
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 50.4,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-3-opus",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-3-opus"
    },
    {
      "name": "Mistral Large",
      "slug": "mistral-large",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-02-26",
      "releaseDateLabel": "Feb 26 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-large",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-large"
    },
    {
      "name": "Mistral Small",
      "slug": "mistral-small",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2024-02-26",
      "releaseDateLabel": "Feb 26 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-small",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-small"
    },
    {
      "name": "Gemma 1",
      "slug": "gemma-1",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-02-21",
      "releaseDateLabel": "Feb 21 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemma-1",
      "markdown": "https://aireleasetracker.com/md/model/google/gemma-1"
    },
    {
      "name": "Gemini 1.5 Pro",
      "slug": "gemini-1.5-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-02-15",
      "releaseDateLabel": "Feb 15 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.5-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.5-pro"
    },
    {
      "name": "Gemini 1.0 Ultra",
      "slug": "gemini-1.0-ultra",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2024-02-08",
      "releaseDateLabel": "Feb 8 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.0-ultra",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.0-ultra"
    },
    {
      "name": "DeepSeek-Math",
      "slug": "deepseek-math",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-02-05",
      "releaseDateLabel": "Feb 5 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-math",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-math"
    },
    {
      "name": "Qwen1.5",
      "slug": "qwen1.5",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2024-02-04",
      "releaseDateLabel": "Feb 4 2024",
      "access": "open-weight",
      "type": null,
      "parameters": "72B",
      "contextWindow": "32k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen1.5",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen1.5"
    },
    {
      "name": "Code Llama 70B",
      "slug": "code-llama-70b",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2024-01-29",
      "releaseDateLabel": "Jan 29 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/code-llama-70b",
      "markdown": "https://aireleasetracker.com/md/model/meta/code-llama-70b"
    },
    {
      "name": "GLM-4",
      "slug": "glm-4",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2024-01-16",
      "releaseDateLabel": "Jan 16 2024",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/zai/glm-4",
      "markdown": "https://aireleasetracker.com/md/model/zai/glm-4"
    },
    {
      "name": "DeepSeek-MoE",
      "slug": "deepseek-moe",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2024-01-11",
      "releaseDateLabel": "Jan 11 2024",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-moe",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-moe"
    },
    {
      "name": "Gemini 1.0 Pro",
      "slug": "gemini-1.0-pro",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2023-12-13",
      "releaseDateLabel": "Dec 13 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.0-pro",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.0-pro"
    },
    {
      "name": "Mistral Medium",
      "slug": "mistral-medium",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2023-12-11",
      "releaseDateLabel": "Dec 11 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-medium",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-medium"
    },
    {
      "name": "Mixtral 8×7B",
      "slug": "mixtral-8×7b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2023-12-08",
      "releaseDateLabel": "Dec 8 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mixtral-8×7b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mixtral-8×7b"
    },
    {
      "name": "Gemini 1.0 Nano",
      "slug": "gemini-1.0-nano",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2023-12-06",
      "releaseDateLabel": "Dec 6 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/gemini-1.0-nano",
      "markdown": "https://aireleasetracker.com/md/model/google/gemini-1.0-nano"
    },
    {
      "name": "Qwen-72B",
      "slug": "qwen-72b",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2023-11-30",
      "releaseDateLabel": "Nov 30 2023",
      "access": "open-weight",
      "type": null,
      "parameters": "72B",
      "contextWindow": "32k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen-72b",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen-72b"
    },
    {
      "name": "DeepSeek-LLM",
      "slug": "deepseek-llm",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2023-11-29",
      "releaseDateLabel": "Nov 29 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-llm",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-llm"
    },
    {
      "name": "Claude 2.1",
      "slug": "claude-2.1",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2023-11-21",
      "releaseDateLabel": "Nov 21 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "200k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/anthropic/claude-2.1",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-2.1"
    },
    {
      "name": "GPT-4 Turbo",
      "slug": "gpt-4-turbo",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2023-11-06",
      "releaseDateLabel": "Nov 6 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "128k",
      "benchmarks": {
        "gpqa": {
          "name": "GPQA Diamond",
          "value": 42.5,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-4-turbo",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4-turbo"
    },
    {
      "name": "Grok‑1",
      "slug": "grok‑1",
      "company": "xai",
      "companyName": "SpaceXAI",
      "releaseDate": "2023-11-03",
      "releaseDateLabel": "Nov 3 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/xai/grok‑1",
      "markdown": "https://aireleasetracker.com/md/model/xai/grok‑1"
    },
    {
      "name": "DeepSeek Coder",
      "slug": "deepseek-coder",
      "company": "deepseek",
      "companyName": "DeepSeek",
      "releaseDate": "2023-11-02",
      "releaseDateLabel": "Nov 2 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/deepseek/deepseek-coder",
      "markdown": "https://aireleasetracker.com/md/model/deepseek/deepseek-coder"
    },
    {
      "name": "Kimi Chat",
      "slug": "kimi-chat",
      "company": "moonshot",
      "companyName": "Moonshot AI",
      "releaseDate": "2023-10-09",
      "releaseDateLabel": "Oct 9 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/moonshot/kimi-chat",
      "markdown": "https://aireleasetracker.com/md/model/moonshot/kimi-chat"
    },
    {
      "name": "Mistral 7B",
      "slug": "mistral-7b",
      "company": "mistral",
      "companyName": "Mistral",
      "releaseDate": "2023-09-27",
      "releaseDateLabel": "Sep 27 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/mistral/mistral-7b",
      "markdown": "https://aireleasetracker.com/md/model/mistral/mistral-7b"
    },
    {
      "name": "Code Llama",
      "slug": "code-llama",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2023-08-24",
      "releaseDateLabel": "Aug 24 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/code-llama",
      "markdown": "https://aireleasetracker.com/md/model/meta/code-llama"
    },
    {
      "name": "Claude Instant 1.2",
      "slug": "claude-instant-1.2",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2023-08-09",
      "releaseDateLabel": "Aug 9 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "100k",
      "benchmarks": {
        "humaneval": {
          "name": "HumanEval",
          "value": 58.7,
          "unit": "%"
        },
        "gsm8k": {
          "name": "GSM8K",
          "value": 86.7,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-instant-1.2",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-instant-1.2"
    },
    {
      "name": "Qwen-7B",
      "slug": "qwen-7b",
      "company": "qwen",
      "companyName": "Qwen",
      "releaseDate": "2023-08-03",
      "releaseDateLabel": "Aug 3 2023",
      "access": "open-weight",
      "type": null,
      "parameters": "7B",
      "contextWindow": "2k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/qwen/qwen-7b",
      "markdown": "https://aireleasetracker.com/md/model/qwen/qwen-7b"
    },
    {
      "name": "LLaMA 2",
      "slug": "llama-2",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2023-07-18",
      "releaseDateLabel": "Jul 18 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/llama-2",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-2"
    },
    {
      "name": "Claude 2",
      "slug": "claude-2",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2023-07-11",
      "releaseDateLabel": "Jul 11 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "100k",
      "benchmarks": {
        "humaneval": {
          "name": "HumanEval",
          "value": 71.2,
          "unit": "%"
        },
        "mmlu": {
          "name": "MMLU",
          "value": 78.5,
          "unit": "%"
        },
        "gsm8k": {
          "name": "GSM8K",
          "value": 88,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/anthropic/claude-2",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-2"
    },
    {
      "name": "Bard",
      "slug": "bard",
      "company": "google",
      "companyName": "Google",
      "releaseDate": "2023-03-21",
      "releaseDateLabel": "Mar 21 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/google/bard",
      "markdown": "https://aireleasetracker.com/md/model/google/bard"
    },
    {
      "name": "Claude 1",
      "slug": "claude-1",
      "company": "anthropic",
      "companyName": "Anthropic",
      "releaseDate": "2023-03-14",
      "releaseDateLabel": "Mar 14 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "9k",
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/anthropic/claude-1",
      "markdown": "https://aireleasetracker.com/md/model/anthropic/claude-1"
    },
    {
      "name": "GPT-4",
      "slug": "gpt-4",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2023-03-14",
      "releaseDateLabel": "Mar 14 2023",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": "8k",
      "benchmarks": {
        "humaneval": {
          "name": "HumanEval",
          "value": 67,
          "unit": "%"
        },
        "mmlu": {
          "name": "MMLU",
          "value": 86.4,
          "unit": "%"
        },
        "gsm8k": {
          "name": "GSM8K",
          "value": 92,
          "unit": "%"
        }
      },
      "url": "https://aireleasetracker.com/model/openai/gpt-4",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-4"
    },
    {
      "name": "ChatGLM-6B",
      "slug": "chatglm-6b",
      "company": "zai",
      "companyName": "Z.ai",
      "releaseDate": "2023-03-14",
      "releaseDateLabel": "Mar 14 2023",
      "access": "open-weight",
      "type": null,
      "parameters": "6B",
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/zai/chatglm-6b",
      "markdown": "https://aireleasetracker.com/md/model/zai/chatglm-6b"
    },
    {
      "name": "LLaMA 1",
      "slug": "llama-1",
      "company": "meta",
      "companyName": "Meta",
      "releaseDate": "2023-02-24",
      "releaseDateLabel": "Feb 24 2023",
      "access": "open-weight",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/meta/llama-1",
      "markdown": "https://aireleasetracker.com/md/model/meta/llama-1"
    },
    {
      "name": "GPT-3.5",
      "slug": "gpt-3.5",
      "company": "openai",
      "companyName": "OpenAI",
      "releaseDate": "2022-11-30",
      "releaseDateLabel": "Nov 30 2022",
      "access": "proprietary",
      "type": null,
      "parameters": null,
      "contextWindow": null,
      "benchmarks": {},
      "url": "https://aireleasetracker.com/model/openai/gpt-3.5",
      "markdown": "https://aireleasetracker.com/md/model/openai/gpt-3.5"
    }
  ]
}