Gemini 3.7 FlashvsKimi K3
Gemini 3.7 Flash | Kimi K3 | |
|---|---|---|
| Specifications | ||
Parameters | — | 2.8T |
Context window | 1M | 1M |
| Benchmarks | ||
Nonsense detection BullshitBench v2Given 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. | — | 73% |
Agentic coding DeepSWE 1.1Artificial 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. | 65.3% | 69%Best |
Agentic coding DeepSWE 1.0Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. Higher is better. | — | 67.5% |
Agentic coding FrontierCode v1.1 (Main) · main splitA set of very hard, frontier-difficulty coding tasks an AI agent has to complete end to end. The score is the share of tasks in the main split it solves. Higher is better. | 43.6% | — |
Next.js coding Next.js EvalsVercel'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. | — | 92% |
Supabase coding Supabase Evals · with skillsSupabase'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. | — | 86.4% |
Supabase coding Supabase Evals · no skillsThe 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. | — | 90.9% |
Agentic terminal coding Terminal-Bench 3.0command-line task completion (v3.0, much harder task set) | 14.9% | — |
Agentic terminal coding Terminal-Bench 2.1Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 85.8% | 88.3%Best |
Multi-step tool use MCP AtlasCan the AI chain together many tools and steps to complete one bigger task, rather than doing just a single thing? Higher is better. | — | 84.2% |
Professional tool use JobBenchTests the AI on professional workplace tasks that require using real work tools — the kind of multi-step jobs an office worker handles. Higher is better. | — | 52.9% |
Personal tool use Toolathlon-VerifiedTests how well the AI uses everyday personal tools and apps to get things done — a human-checked version of Toolathlon. Higher is better. | — | 73.2% |
Web browsing BrowseCompCan the AI browse the web and track down hard-to-find answers? Higher is better. | — | 91.2% |
Multidisciplinary reasoning Humanity's Last Exam · no toolsHumanity'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. | — | 43.5% |
Multidisciplinary reasoning Humanity's Last Exam · with toolsHumanity'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. | — | 56% |
Multidisciplinary reasoning Humanity's Last Exam (Verified)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. | 53.6% | — |
Biology BioMysteryBench · hardReal 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. | 43.5% | — |
Biology BioMysteryBench · human solvedReal biology puzzles that human experts have managed to crack — can the AI reach the same answers? Higher is better. | 87.1% | — |
Biology LAB-Bench 2Everyday 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. | 82.1% | — |
Science GPQA DiamondGraduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 93.5% |
Agentic computer use OSWorld 2.0Can 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. | 38.1% | — |
Agentic computer use Agent's Last ExamA hard set of desktop and operating-system tasks an AI agent has to finish by looking at the screen and working the machine itself. The score is the share it passes outright — partial credit does not count. Higher is better. | 26.3% | — |
Business workflows AutomationBenchTests whether the AI can run real multi-step business workflows — the kind of end-to-end office processes companies want to automate — from start to finish. Higher is better. | 30.4% | — |
Agentic legal work Harvey's Legal Agent BenchmarkHarvey'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. | 90.7% | — |
Overall intelligence AA Intelligence IndexArtificial Analysis composite intelligence index across evals | 56 | — |
Knowledge work GDPval-AA v2economically valuable knowledge work (v2, re-based Elo) | 1525 | 1668Best |
Chart reasoning CharXiv ReasoningCan the AI read and reason about complex charts and figures, not just text? Higher is better. | 84.5% | 84.8%Best |
Document comprehension GDP.PDFReal 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% | — |
Video understanding LVBenchCan 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. | 85.4% | — |
Multimodal reasoning MMMU-ProA tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better. | — | 81.6% |
Long context MRCR v2 (8-needle) · 128k averageTests whether the AI can find specific details buried inside a very long document (around 128k tokens — roughly a long book). Higher is better. | 97% | — |
Long context MRCR v2 (8-needle) · 1M pointwiseTests whether the AI can find specific details buried inside an enormous document (around 1 million tokens — many books). Higher is better. | 62.5% | — |
Community preference Arena Elo (Text)Real people chat with two anonymous AIs side by side and vote for the answer they prefer. Votes become a chess-style Elo rating on arena.ai — it measures which AI people actually like, not test scores. Higher is better. | — | 1486 |
Community preference (code) Arena Elo (Code)Like the text arena, but people vote on which AI writes better code. The votes become a chess-style Elo rating on arena.ai. Higher is better. | 1588 | 1679Best |
| Overview | ||
| Company | Moonshot AI | |
| Release date | Aug 13 2026 | Jul 16 2026 |
| Access | Proprietary | Open Weight |
Which is better: Gemini 3.7 Flash or Kimi K3?
Kimi K3 leads Gemini 3.7 Flash on 5 of the 5 benchmarks they both report. Kimi K3 shipped 28 days before Gemini 3.7 Flash, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (Gemini 3.7 Flash) vs 1M (Kimi K3). Gemini 3.7 Flash is proprietary, while Kimi K3 is open weight.
On DeepSWE 1.1, Kimi K3 leads at 69% vs Gemini 3.7 Flash at 65.3%. On Terminal-Bench 2.1, Kimi K3 leads at 88.3% vs Gemini 3.7 Flash at 85.8%. On GDPval-AA v2, Kimi K3 leads at 1668 vs Gemini 3.7 Flash at 1525. On CharXiv Reasoning, Kimi K3 leads at 84.8% vs Gemini 3.7 Flash at 84.5%. On Arena Elo (Code), Kimi K3 leads at 1679 vs Gemini 3.7 Flash at 1588.
Frequently asked questions
Gemini 3.7 Flash was released by Google on Aug 13 2026.
Kimi K3 was released by Moonshot AI on Jul 16 2026.
Kimi K3 leads on DeepSWE 1.1 — Gemini 3.7 Flash 65.3% vs Kimi K3 69%.
Gemini 3.7 Flash has a 1M context window; Kimi K3 has 1M.
Gemini 3.7 Flash is a proprietary model released by Google. Kimi K3 is an open weight model released by Moonshot AI.
Other comparisons
Gemini 3.7 FlashvsClaude Opus 5Kimi K3vsClaude Opus 5Gemini 3.7 FlashvsGPT-5.6-CyberKimi K3vsGPT-5.6-CyberGemini 3.7 FlashvsMuse GlimmerKimi K3vsMuse GlimmerGemini 3.7 FlashvsGrok 4.6Kimi K3vsGrok 4.6Gemini 3.7 FlashvsDeepSeek-V4-Pro-0813Kimi K3vsDeepSeek-V4-Pro-0813Gemini 3.7 FlashvsMistral Medium 3.5Kimi K3vsMistral Medium 3.5