Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 2.8T | — |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 1M | 1.05M |
| API pricingUSD per 1M tokens · lower wins · base tier | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | $2.00 | $8.00 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | $10.00 | $40.00 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | $0.20 | $0.40 |
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.2403Relace | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $6.10Decart | — |
| $468Relace · fp4 | $3,600OpenAI list price | |
| Benchmarks | ||
Terminal-Bench-Science 0.1Agentic scientific computing — The same command-line setup as Terminal-Bench, pointed at scientific work: the AI has to drive research tooling and computational workflows through to a result, rather than administer a machine. Version 0.1 is the first release of the task set, and scores run lower than on the general board. Higher is better. | 38.1% | 77.1% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 94.2% | 97% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 60% | — |
Blocker Deception (OpenAI) · without Codex promptBlocker disclosure — share of coding trials where the model omitted or misrepresented an obstacle without the Codex developer prompt | — | 22.28% |
Blocker Deception (OpenAI) · with Codex promptBlocker disclosure — share of coding trials where the model omitted or misrepresented an obstacle with the Codex developer prompt | — | 8.04% |
Simulated Workplace Tasks (OpenAI)High-severity improper actions — share of scored simulated workplace tasks with a high-severity improper action | — | 0.64% |
DeepSWE 1.1Agentic coding — Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. (Version 1.1 of the test.) Higher is better. | 75.5% | — |
SWE-Marathon v1.1Multi-hour coding — Twenty multi-hour programming jobs — rebuilding a library, cloning a product, training a model, making something faster. A run counts only when every check passes, hidden tests and interface review included. Higher is better. | 51.7% | — |
FrontierCode 2.0Agentic coding — whether a model’s code changes to open-source repositories would be merged by their maintainers (version 2.0); separate from FrontierCode v1.1 scores | — | 44.2% |
BenchCADCAD programming — Can the AI do mechanical design as code? Given a drawing or a description of an industrial part — a gear, a spring, a drill bit — it has to write or edit the parametric CAD program that builds it, and the program is run to check the shape really comes out right. Higher is better. | 92.7% | — |
BenchCAD 2.1 CoreCAD programming — partial-credit score for modeling hardware designs in the held-out Core evaluation | — | 82.5% |
SecureSWESecure code generation — OpenAI’s internal evaluation of writing code that is both functional and secure | — | 65.7% |
PatchBenchSecurity patching — security patches that fix vulnerabilities while preserving intended behavior | — | 79.7% |
Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. | 65.7% | — |
APEX-AgentsExpert agentic work — expert-level agentic work tasks (AI Productivity Index) | 67.2% | — |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 92.1% | — |
ExploitBenchCybersecurity — A 'capability ladder' for security research, built by CMU researchers: the AI is given known bugs in Chrome's V8 engine and scored on how far it gets toward a working exploit inside a research sandbox — from understanding the patch to triggering a crash. Higher is better. | 74.7% | — |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | 51.1% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 70% | — |
Vals Science Index v1Scientific reasoning — Vals AI’s composite accuracy across three scientific reasoning evaluations | — | 68.4% |
Agents’ Last Exam v1.1Agentic computer use — averaged rubric score on the 122-task v1.1 suite of long-horizon professional workflows | — | 78.3% |
AutomationBench 1.0.6Business workflows — end-to-end business workflow automation on the version 1.0.6 task set | — | 52.4% |
LifeValEveryday tasks — OpenAI’s rubric score for completing practical everyday tasks tailored to a person’s needs | — | 86.1% |
AA-AnalystAgent (pass^5)Financial research — share of quantitative questions about real spreadsheets and reports answered correctly on all five independent attempts | — | 62.5% |
Harvey LAB-AA v1.1Agentic legal work — full-credit rate on legal assignments, requiring every grading criterion to be met without material hallucinations | — | 15% |
HealthBench Professional · OpenAI grading, rawHealth — unadjusted physician-rubric score using the official OpenAI grading procedure | — | 66% |
HealthBench Professional · OpenAI grading, length-adjustedHealth — official OpenAI grading with the response-length penalty applied | — | 66.3% |
MentalHealthBenchMental health — overall clinician-rubric score on mental health conversations | — | 53.3% |
AA Intelligence Index v5.0.1Overall intelligence — Artificial Analysis’s version 5.0.1 composite across nine evaluations of knowledge work, coding, science, and general knowledge | — | 62.1 |
GDPval-AA v2.1Knowledge work — economically valuable knowledge work (v2.1, Crowd-BT Elo fit) | 1846 | — |
AA-Briefcase v1.1Knowledge work — Artificial Analysis agentic office-work eval (Elo, v1.1 rating fit) | 1678 | — |
GDPval v2 (OpenAI Internal)Knowledge work — OpenAI’s internal score on multistep professional assignments drawn from consulting, finance, and design projects | — | 50.4% |
CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better. | 89% | — |
CharXiv Reasoning · with toolsChart reasoning — The same chart-and-figure reasoning test, run with the AI allowed to use tools — writing code to inspect the image, for instance — rather than reading the chart unaided. Scores run higher than the unaided version, so read the two as separate tests. Higher is better. | 93.4% | — |
Chartography · with toolsChart tasks — A chart-centred test run with tools available to the AI, reported separately from the chart-reading benchmarks above it. Higher is better. | 90.2% | — |
OfficeQA ProDocument Q&A — Questions about office documents, where answering depends on reading the page as a document — layout, tables and figures included — rather than as loose text. Higher is better. | 67.4% | — |
GDP.PDFDocument comprehension — Real professional PDFs — filings, reports, technical documents — with questions an expert in that field would ask. Tests whether the AI reads the page as a document, layout and figures included, rather than as loose text. Higher is better. | 34.6% | — |
MMMU-ProMultimodal reasoning — A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better. | 84.1% | — |
AA-3DCommunity preference (interactive 3D) — blind human preferences between paired interactive 3D experiences built from the same prompt | — | 1318 |
| Overview | ||
| Company | Moonshot AI | OpenAI |
| Release date | Oct 23 2026 | Nov 17 2026 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
Kimi K3.1 ProvsClaude Fable 5.2GPT-6.2 AstravsClaude Fable 5.2Kimi K3.1 ProvsGemini 4 HydrogenGPT-6.2 AstravsGemini 4 HydrogenKimi K3.1 ProvsMuse EmberGPT-6.2 AstravsMuse EmberKimi K3.1 ProvsGrok 4.7GPT-6.2 AstravsGrok 4.7Kimi K3.1 ProvsDeepSeek-V4.1-Flash-1017GPT-6.2 AstravsDeepSeek-V4.1-Flash-1017Kimi K3.1 ProvsMistral Large 4GPT-6.2 AstravsMistral Large 4Frequently asked questions
GPT-6.2 Astra leads Kimi K3.1 Pro on 2 of the 2 benchmarks they both report (Terminal-Bench-Science 0.1, GPQA Diamond). Kimi K3.1 Pro is cheaper on both input and output: $2.00 vs $8.00 per million input tokens, and $10.00 vs $40.00 per million output tokens. Figures are base-tier rates. Kimi K3.1 Pro shipped 25 days before GPT-6.2 Astra, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (Kimi K3.1 Pro) vs 1.05M (GPT-6.2 Astra). Kimi K3.1 Pro is open weight, while GPT-6.2 Astra is closed.
On Terminal-Bench-Science 0.1, GPT-6.2 Astra leads at 77.1% vs Kimi K3.1 Pro at 38.1%. On GPQA Diamond, GPT-6.2 Astra leads at 97% vs Kimi K3.1 Pro at 94.2%.