TestsBenchmarks · Agentic
Terminal-Bench 2.1
Share of 89 containerized command-line tasks an agent completes; a fixed revision of Terminal-Bench 2.0, now archived and near ceiling.
% solvedHigher is betterToo easy now: the top models all score near perfectSaturatedThe test's websiteOfficial page ↗
Measured by independent testersIndependent Reported by the makerVendor-reported
| Thinking levelSetting | |||||||
|---|---|---|---|---|---|---|---|
| 1 | GPT-5.6 Sol | OpenAI | 91.9% | DefaultGPT-5.6 Sol Ultra (4 parallel agents) | Maker's own figureVendor-reportedOpenAI ↗ | Codex | 9 Jul 2026 |
| 2 | Claude Fable 5.1 | Anthropic | 91.4% | MaxClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 3 | DeepSeek V4.1 Flash | DeepSeek | 90.6% | Maxreasoning_effort=100 (max) | Maker's own figureVendor-reportedDeepSeek ↗ | DeepSeek Harness (Minimal) | 10 Sep 2026 |
| 4 | GPT-6 Astra | OpenAI | 89.9% | HighGPT-6 Astra (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 5 | Gemini 3.8 Flash | Google (Gemini / DeepMind) | 89.4% | Defaultdefault settings (API default thinking_level=medium) | Maker's own figureVendor-reportedGoogle ↗ | Terminus 2 | 2 Sep 2026 |
| 6 | Claude Opus 5 | Anthropic | 89.1% | MaxClaude Opus 5 (Adaptive Reasoning, Max Effort) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 7 | Qwen3.8 Max (0902) | Qwen (Alibaba) | 88.8% | DefaultQwen3.8 Max (0902) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 8 | Grok 4.6 | SpaceXAI (formerly xAI) | 88.4% | HighGrok 4.6 (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 9 | GPT-5.6 Terra | OpenAI | 88.0% | MaxGPT-5.6 Terra (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 10 | Claude Opus 5.5 | Anthropic | 87.6% | Higheffort=high | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 11 | Qwen3.8 Max (0803) | Qwen (Alibaba) | 86.6% | Defaultthinking (default reasoning_effort=xhigh; eval setting not stated) | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Claude Code | 3 Aug 2026 |
| 12 | Qwen3.8-Flash-Next | Qwen (Alibaba) | 86.1% | DefaultQwen3.8-Flash-Next | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 13 | Gemini 3.7 Flash | Google (Gemini / DeepMind) | 85.8% | HighGemini 3.7 Flash (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 14 | Muse Spark 1.3 | Meta (Meta Superintelligence Labs, Muse) | 85.4% | Extra highMuse Spark 1.3 (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 15 | Kimi K3 | Moonshot AI (Kimi) | 85.0% | MaxKimi K3 (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 16 | Claude Opus 4.8 | Anthropic | 84.6% | MaxClaude Opus 4.8 (Adaptive Reasoning, Max Effort) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 17 | Claude Fable 5 | Anthropic | 84.6% | MaxClaude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 18 | GLM-5.3-Flash | Z.ai (Zhipu) | 84.3% | DefaultGLM 5.3 Flash | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 19 | GPT-5.5 | OpenAI | 84.3% | Extra highGPT-5.5 (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 20 | GLM-5.3 | Z.ai (Zhipu) | 83.9% | MaxGLM-5.3 (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 21 | GPT-6 Sol | OpenAI | 83.2% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 22 | Claude Sonnet 5.5 | Anthropic | 83.2% | Higheffort=high | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 23 | Claude Opus 4.7 | Anthropic | 83.1% | MaxClaude Opus 4.7 (Adaptive Reasoning, Max Effort) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 24 | Qwen3.8 2.4T A95B (open weights) | Qwen (Alibaba) | 82.0% | DefaultQwen3.8 2.4T A95B | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 25 | Grok 4.5 | SpaceXAI (formerly xAI) | 81.6% | HighGrok 4.5 (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 26 | GLM-5.2 | Z.ai (Zhipu) | 81.0% | Defaultthinking (effort not stated) | Maker's own figureVendor-reportedZ.ai ↗ | Terminus 2 | 16 Jun 2026 |
| 27 | GPT-5.6 Luna | OpenAI | 80.9% | MaxGPT-5.6 Luna (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 28 | Claude Sonnet 5 | Anthropic | 80.5% | MaxClaude Sonnet 5 (Adaptive Reasoning, Max Effort) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 29 | Muse Spark 1.2 | Meta (Meta Superintelligence Labs, Muse) | 80.1% | Extra highMuse Spark 1.2 (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 30 | Qwen3.8 27B | Qwen (Alibaba) | 79.8% | Extra highQwen3.8 27B (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 31 | DeepSeek V4 Pro (0813) | DeepSeek | 78.7% | MaxDeepSeek V4 Pro 0813 (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 32 | DeepSeek V4 Flash (0731) | DeepSeek | 78.7% | MaxDeepSeek V4 Flash 0731 (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 33 | GPT-5.4 | OpenAI | 78.3% | Extra highGPT-5.4 (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 34 | Gemini 3.6 Flash | Google (Gemini / DeepMind) | 78.0% | Defaultdefault settings (API default thinking_level=medium) | Maker's own figureVendor-reportedGoogle ↗ | Terminus 2 | 21 Jul 2026 |
| 35 | Muse Spark 1.1 | Meta (Meta Superintelligence Labs, Muse) | 77.9% | Extra highMuse Spark 1.1 (Xhigh) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 36 | MiMo-V2.6-Flash | Xiaomi MiMo | 76.4% | Default | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 37 | Gemini 3.5 Flash | Google (Gemini / DeepMind) | 76.2% | Defaultdefault settings (API default thinking_level=medium) | Maker's own figureVendor-reportedGoogle ↗ | Terminus 2 | 19 May 2026 |
| 38 | Qwen3.7 Max | Qwen (Alibaba) | 74.5% | Defaultthinking (setting not stated) | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Claude Code | 3 Aug 2026 |
| 39 | DeepSeek V4 Flash Vision Exp | DeepSeek | 74.2% | MaxDeepSeek V4 Flash Vision (Max) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 40 | Gemini 3 Pro | Google (Gemini / DeepMind) | 73.9% | Default | Independent testIndependentSnorkel AI / Terminal-Bench ↗ | Terminus 2 | 1 Oct 2026 |
| 41 | Grok 4.7 | SpaceXAI (formerly xAI) | 73.4% | Extra highreasoning_effort=xhigh | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 42 | GPT-6 Luna | OpenAI | 73.0% | Maxreasoning_effort=max | Independent testIndependentVals.ai ↗ | Terminus 2 | 27 Sep 2026 |
| 43 | DeepSeek V4 Pro (Preview, 0423) | DeepSeek | 72.1% | Defaultsetting not stated for preview columns | Maker's own figureVendor-reportedDeepSeek ↗ | — | 13 Aug 2026 |
| 44 | Muse Spark | Meta (Meta Superintelligence Labs, Muse) | 67.3% | Defaulteffort not stated | Maker's own figureVendor-reportedMeta ↗ | — | 9 Jul 2026 |
| 45 | Claude Sonnet 4.6 | Anthropic | 67.0% | Highadaptive thinking, effort=high | Maker's own figureVendor-reportedAnthropic ↗ | mini-SWE-agent (GKE) | 30 Jun 2026 |
| 46 | Gemini 3.1 Pro (Preview) | Google (Gemini / DeepMind) | 65.8% | Default | Independent testIndependentSnorkel AI / Terminal-Bench ↗ | Gemini CLI | 1 Oct 2026 |
| 47 | MiniMax M3 | MiniMax | 65.2% | DefaultMiniMax-M3 | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 48 | Hy3 | Tencent Hunyuan | 64.4% | DefaultHy3 | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 49 | Qwen3.7 Plus | Qwen (Alibaba) | 64.0% | Defaultsetting not stated | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 | 14 Aug 2026 |
| 50 | Qwen3.6 27B | Qwen (Alibaba) | 63.4% | Defaultsetting not stated | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 | 14 Aug 2026 |
| 51 | DeepSeek V4 Flash (Preview, 0423) | DeepSeek | 61.8% | Defaultsetting not stated for preview columns | Maker's own figureVendor-reportedDeepSeek ↗ | — | 13 Aug 2026 |
| 52 | GLM-5.1 | Z.ai (Zhipu) | 58.7% | Default | Independent testIndependentSnorkel AI / Terminal-Bench ↗ | Claude Code | 1 Oct 2026 |
| 53 | Gemini 3.5 Flash-Lite | Google (Gemini / DeepMind) | 54.0% | Highhigh thinking | Maker's own figureVendor-reportedGoogle ↗ | Terminus 2 | 21 Jul 2026 |
| 54 | Nemotron 3 Ultra | NVIDIA | 53.9% | DefaultNemotron 3 Ultra 550B A55B (Reasoning) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 55 | Muse Glimmer 30B | Meta (Meta Superintelligence Labs, Muse) | 51.7% | HighMuse Glimmer (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 56 | MiniMax M2.7 | MiniMax | 51.1% | Defaultsetting not stated | Maker's own figureVendor-reportedMiniMax ↗ | Terminus 2 | 1 Jun 2026 |
| 57 | Mistral Medium 3.5 | Mistral AI | 50.6% | DefaultMistral Medium 3.5 | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 58 | Qwen3.6 35B A3B | Qwen (Alibaba) | 44.9% | DefaultQwen3.6 35B A3B (Reasoning) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 59 | Gemma 4 31B IT | Google (Gemini / DeepMind) | 43.5% | DefaultGemma 4 31B (Reasoning) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 60 | Nemotron 3 Super 120B A12B | NVIDIA | 38.6% | DefaultNemotron 3 Super 120B A12B (Reasoning) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 61 | Gemini 3.1 Flash-Lite | Google (Gemini / DeepMind) | 31.0% | Highhigh thinking | Maker's own figureVendor-reportedGoogle ↗ | Terminus 2 | 21 Jul 2026 |
| 62 | gpt-oss-120b | OpenAI | 26.2% | Highgpt-oss-120b (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 63 | Nemotron 3.5 Lightning 30B-A3B | NVIDIA | 24.3% | DefaultNemotron 3.5 Lightning | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 64 | Command A+ | Cohere | 22.9% | DefaultCommand A+ | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 65 | Mistral Small 4 | Mistral AI | 21.0% | HighMistral Small 4 (Reasoning) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 66 | Granite 4.2 8B | IBM Granite | 18.4% | DefaultGranite 4.2 8B | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 67 | gpt-oss-20b | OpenAI | 13.9% | Highgpt-oss-20b (High) | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |
| 68 | Ministral 3 14B | Mistral AI | 9.7% | DefaultMinistral 3 14B | Independent testIndependentArtificial Analysis ↗ | — | 1 Oct 2026 |