Compare AI models
| Specifications | ||
Context windowHow much text the model can hold in mind at once — your question, any documents you attach, the conversation so far, and its own reply. Go past it and the earliest part falls out of view. | 1M | — |
| 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 |
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. | — | $6.00 |
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. | — | $0.50 |
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.375Google | $2.00xAI |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | $1.875Google | $6.00xAI |
| Benchmarks | ||
CheatBenchCheating rate — Measures how often AI agents try to cheat on difficult assignments, such as reading hidden answers, copying work or manipulating grading. The overall score gives equal weight to ten categories; the sycophancy category measures how far an agent shifts its beliefs toward a user's stated views. Scores describe each model in its tested agent setup, and task success is measured separately. Lower is better. | 75.2% | 79.1% |
CursorBench 4.0Agentic coding — Cursor's own test of coding agents on ambiguous, multi-file tasks taken from real Cursor sessions — editing, refactoring, investigating a codebase, understanding what the user meant, managing jobs and following a design. Cursor runs each model at several reasoning efforts; each release here carries the score of its best listed effort. Scores aren't comparable with earlier CursorBench versions. Higher is better. | 39.6% | 41.4% |
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. | 71% | 65.9% |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | 90% | 71% |
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. | 19.1% | 20.3% |
ARC-AGI-2Abstract reasoning — Puzzle-style tests of abstract reasoning and pattern-finding — the kind of thing people find easy but AIs often struggle with. Higher is better. | 89.2% | 67.1% |
Harvey's Legal Agent BenchmarkAgentic legal work — Harvey's test of whether an AI agent can complete real legal work — drafting and reviewing documents, working with spreadsheets and presentations, and navigating files the way a lawyer's assistant would. Higher is better. | 10% | 15.8% |
GDPval-AA v2Knowledge work — economically valuable knowledge work (v2, re-based Elo) | 1545 | 1753 |
threejsevalCommunity preference (Three.js) — Every model gets the same prompt — "the Eiffel Tower", "a glass fishbowl", "a robot arm picking toys into a box" — and builds a 3D scene in Three.js. Real people then see two scenes side by side, names hidden, and vote for the one they prefer. The votes become a chess-style Elo rating on threejseval.com, averaged across all the prompts. It measures whether the scene looks and moves right to a human eye, not whether the code passes a test. Higher is better. | 1557 | 1450 |
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. | — | 66% |
FrontierSWE V2Ultra-long-horizon coding — Engineering problems that would occupy a person for days: systems implementation, performance work, scientific computing and AI research, with up to twenty hours per task. Partial credit is awarded, because finishing one outright is still rare. Higher is better. | — | 25.3% |
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. | — | 31.9% |
FrontierCode v1.1 (Extended) · extended splitAgentic coding — frontier-difficulty agentic coding tasks (v1.1, extended split) | — | 61.3% |
APEX-SWEExpert software engineering — expert-level software-engineering tasks (AI Productivity Index) | — | 56.4% |
Supabase Evals · with skillsSupabase coding — Supabase's own open benchmark: a coding agent is dropped into a real Supabase project and asked to do real work — set up a schema, fix a broken security policy, debug an Edge Function — and every run is checked against a live Supabase stack. This is the headline number, where the agent has Supabase's own skills loaded, as most people building on Supabase would. The score is the share of scenarios it got right. Higher is better. | — | 84.1% |
Supabase Evals · no skillsSupabase coding — The same Supabase scenarios, but with none of Supabase's skills loaded — so it measures what the model already knows about building on Supabase, rather than how well it follows Supabase's supplied instructions. Higher is better. | — | 81.2% |
CADGenBenchCAD generation — Can the AI turn a written design brief into a real CAD model? What it produces has to run as a program and come out the right shape. Higher is better. | — | 40.9% |
Terminal-Bench 3.0Agentic terminal coding — command-line task completion (v3.0, much harder task set) | — | 26% |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 89.4% | — |
APEX-AgentsExpert agentic work — expert-level agentic work tasks (AI Productivity Index) | — | 57.5% |
CyberGymCybersecurity — Tests the AI on cybersecurity challenges — finding and exploiting software weaknesses inside a safe sandbox. Higher is better. | — | 79.7% |
CVE-BenchCybersecurity — Can the AI exploit real, publicly known security holes in web applications, in a sandbox built to resemble a live service? A measure of raw capability: it is run without the safeguards a deployed model carries. Higher is better. | — | 39.8% |
CathedralBenchCybersecurity — An outside red team's test: chaining several exploits together to get through a hardened system, scored on the hardest tasks only and run without safeguards, to show what the underlying model can do. Higher is better. | — | 25% |
Humanity's Last Exam (Verified)Multidisciplinary reasoning — The re-checked edition of Humanity's Last Exam: the same extremely hard expert questions, minus the ones found to be flawed or wrongly answered. Scores on it run lower than on the original exam, so read the two as separate tests rather than a before-and-after. Higher is better. | 54.9% | — |
BioMysteryBench · hardBiology — Real unsolved-style biology puzzles — the AI has to reason its way to an answer the way a research biologist would. The “hard” split contains the toughest cases. Higher is better. | 56.5% | — |
BioMysteryBench · human solvedBiology — Real biology puzzles that human experts have managed to crack — can the AI reach the same answers? Higher is better. | 88.8% | — |
LatchBio Capabilities v1.0Biology — Can the AI do a working biologist's analysis — opening real experimental files and reaching a defensible conclusion across genomics, drug discovery and pathogen surveillance? Averaged over eleven separate tests. Higher is better. | — | 43.3% |
LAB-Bench 2Biology — Everyday tasks from a working biology lab — reading protocols, interpreting figures and sequence data, and answering the practical questions a researcher hits at the bench. Higher is better. | 86.2% | — |
EEBenchElectrical engineering — Can the AI design electronics? It is asked for circuits and chip layouts, and the designs are checked for whether they would physically work — not whether the answer reads well. Higher is better. | — | 60% |
OSWorld 2.0Agentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Version 2.0 is a harder, refreshed task set. Higher is better. | 59% | — |
Finance Agent v2Agentic financial analysis — Tests the AI on real financial-analysis work, like digging through reports and making sound decisions. Higher is better. | 61.4% | — |
HealthBench ProfessionalHealth — Realistic health conversations graded against detailed rubrics written by physicians — can the AI respond the way a careful medical professional would? Higher is better. | — | 48.5% |
AA Intelligence IndexOverall intelligence — Artificial Analysis composite intelligence index across evals | — | 61 |
AA-BriefcaseKnowledge work — Artificial Analysis agentic office-work eval (Elo) | — | 1577 |
CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better. | 86.2% | — |
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. | 35% | — |
LVBenchVideo understanding — Can the AI follow a very long video — up to an hour — and answer questions that need details from far apart in it? Higher is better. | 87.1% | — |
LVBench · agenticVideo understanding — The same long-video test, run with the AI free to navigate the video itself — skipping around, replaying sections — rather than being shown it once. Read it as a separate test from the unaided run above. Higher is better. | 87.8% | — |
| Overview | ||
| Company | SpaceXAI | |
| Release date | Sep 2 2026 | Aug 12 2026 |
| Access | Closed | Closed |
| Model details | View model | View model |
Other comparisons
Gemini 3.8 FlashvsClaude Sonnet 5.5Grok 4.6vsClaude Sonnet 5.5Gemini 3.8 FlashvsGPT-6.1 SolGrok 4.6vsGPT-6.1 SolGemini 3.8 FlashvsMuse Spark 1.3Grok 4.6vsMuse Spark 1.3Gemini 3.8 FlashvsDeepSeek-V4.1-FlashGrok 4.6vsDeepSeek-V4.1-FlashGemini 3.8 FlashvsMistral Large 4Grok 4.6vsMistral Large 4Gemini 3.8 FlashvsKimi K3Grok 4.6vsKimi K3Frequently asked questions
Gemini 3.8 Flash leads Grok 4.6 on 5 of the 9 benchmarks they both report. Only Grok 4.6 has a verified first-party API price: $2.00 per million input tokens and $6.00 per million output tokens. No pay-as-you-go API rate is tracked for Gemini 3.8 Flash. Grok 4.6 shipped 21 days before Gemini 3.8 Flash, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On CheatBench, Gemini 3.8 Flash leads at 75.2% vs Grok 4.6 at 79.1%. On CursorBench 4.0, Grok 4.6 leads at 41.4% vs Gemini 3.8 Flash at 39.6%. On DeepSWE 1.1, Gemini 3.8 Flash leads at 71% vs Grok 4.6 at 65.9%. On Next.js Evals, Gemini 3.8 Flash leads at 90% vs Grok 4.6 at 71%. On Terminal-Bench 4.0, Grok 4.6 leads at 20.3% vs Gemini 3.8 Flash at 19.1%. On ARC-AGI-2, Gemini 3.8 Flash leads at 89.2% vs Grok 4.6 at 67.1%. On Harvey's Legal Agent Benchmark, Grok 4.6 leads at 15.8% vs Gemini 3.8 Flash at 10%. On GDPval-AA v2, Grok 4.6 leads at 1753 vs Gemini 3.8 Flash at 1545. On threejseval, Gemini 3.8 Flash leads at 1557 vs Grok 4.6 at 1450.