Terminal-Bench 2.1 — Distinct vintage from 2.0; variants incl. codex-cli/gemini-cli/terminus-2/best-reported-harness
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | GPT-5.6 Sol | OpenAI | 89.5 | 9 | 2026-08-23 |
| 2 | Opus 5 | Anthropic | 89.1 | 5 | 2026-07-25 |
| 3 | Grok 4.6 | SpaceXAI | 88.4 | 4 | 2026-08-22 |
| 4 | Kimi K3 | Moonshot | 88.3 | 6 | 2026-08-23 |
| 5 | GLM-5.3 | Z.ai | 88.2 | 2 | 2026-08-23 |
| 6 | GPT-5.6 Terra | OpenAI | 88 | 7 | 2026-08-23 |
| 7 | Claude Fable 5 | Anthropic | 88 | 6 | 2026-08-23 |
| 8 | Claude Mythos 5 | Anthropic | 88 | 2 | 2026-06-12 |
| 9 | Qwen3.8 Max Preview | Alibaba | 86.6 | 3 | 2026-08-23 |
| 10 | Gemini 3.7 Flash | 85.8 | 6 | 2026-08-24 | |
| 11 | Claude Opus 4.8 | Anthropic | 85 | 10 | 2026-08-15 |
| 12 | GPT-5.6 Luna | OpenAI | 84.7 | 8 | 2026-08-23 |
| 13 | GPT-5.5 | OpenAI | 84.3 | 11 | 2026-08-15 |
| 14 | DeepSeek-V4-Flash-Vision-Exp | DeepSeek | 83.9 | 1 | 2026-08-23 |
| 15 | Grok4.5 | SpaceXAI | 83.3 | 2 | 2026-08-23 |
| 16 | Claude Opus 4.7 | Anthropic | 83.1 | 3 | 2026-07-06 |
| 17 | Muse Spark 1.2 | Meta | 82.9 | 2 | 2026-08-23 |
| 18 | GLM-5.2 Full Open Source | Z.ai | 82.7 | 7 | 2026-08-23 |
| 19 | DeepSeek-V4-Flash | DeepSeek | 82.7 | 11 | 2026-08-23 |
| 20 | Qwen3.8 2.4T A95B | Alibaba | 82 | 1 | 2026-08-14 |
| 21 | Claude Sonnet 5 | Anthropic | 80.5 | 3 | 2026-07-07 |
| 22 | Muse Spark 1.1 | Meta | 80 | 3 | 2026-08-23 |
| 23 | Qwen3.8 27B | Alibaba | 79.8 | 5 | 2026-08-23 |
| 24 | Gemini 3.5 Flash | 78.7 | 6 | 2026-08-24 | |
| 25 | GPT-5.4 | OpenAI | 78.3 | 1 | 2026-06-17 |
| 26 | Gemini 3.6 Flash | 78 | 4 | 2026-08-24 | |
| 27 | Qwen3.7 Max | Alibaba | 75 | 3 | 2026-08-15 |
| 28 | Motif-3 | Motif Technologies | 74.9 | 1 | 2026-08-13 |
| 29 | Gemini 3.1 Pro | 74 | 6 | 2026-08-15 | |
| 30 | DeepSeek-V4-Pro | DeepSeek | 72.1 | 5 | 2026-08-15 |
| 31 | Hy3-preview | Tencent | 71.7 | 2 | 2026-08-23 |
| 32 | Claude Sonnet 4.6 | Anthropic | 71.2 | 3 | 2026-07-07 |
| 33 | Motif-3-Beta | Motif Technologies | 70.8 | 1 | 2026-07-21 |
| 34 | laguna-s-2.1 | poolside | 70.2 | 2 | 2026-08-23 |
| 35 | Nex-N2-Pro | Nex AGI | 67.8 | 2 | 2026-07-07 |
| 36 | Kimi K2.7 Code | Moonshot | 67.4 | 2 | 2026-07-07 |
| 37 | Kimi K2.6 1T A32B | Moonshot | 67.2 | 1 | 2026-06-06 |
| 38 | Agnes 2.5 Pro Alpha | Sapiens AI | 67 | 1 | 2026-07-25 |
| 39 | Claudes | Anthropic | 67 | 1 | 2026-08-12 |
| 40 | MiniMax M3 | MiniMax | 66 | 4 | 2026-08-23 |
| 41 | Kimi K2.6 | Moonshot | 65.9 | 2 | 2026-07-07 |
| 42 | MiMo V2.5 Pro | Xiaomi | 65.2 | 3 | 2026-07-30 |
| 43 | Inkling-Small | Thinking Machines | 64.7 | 3 | 2026-08-23 |
| 44 | Inkling | Thinking Machines | 63.8 | 2 | 2026-07-16 |
| 45 | GLM-5.1 | Z.ai | 63.5 | 2 | 2026-08-15 |
| 46 | Muse Spark | Meta | 62.2 | 2 | 2026-07-07 |
| 47 | MiMo V2 Flash | Xiaomi | 61.8 | 1 | 2026-06-27 |
| 48 | Qwen3.6 Plus | Alibaba | 61.4 | 2 | 2026-07-07 |
| 49 | Qwen3.7 Plus Preview | Alibaba | 61 | 2 | 2026-07-07 |
| 50 | Qwen3.6 27B | Alibaba | 60.7 | 2 | 2026-07-03 |
| 51 | GPT-5.4 nano | OpenAI | 60.7 | 1 | 2026-06-17 |
| 52 | JT-4.1 Flash 236B A21B | China Mobile | 59.6 | 1 | 2026-07-10 |
| 53 | GLM-5.1-744B-A40B | Z.ai | 59.3 | 1 | 2026-06-06 |
| 54 | GPT-5.4 mini | OpenAI | 59.2 | 1 | 2026-06-17 |
| 55 | Gemini 3 Flash Preview | 58 | 1 | 2026-07-06 | |
| 56 | Solar Pro4 | Upstage | 57.3 | 2 | 2026-08-23 |
| 57 | Nemotron 3 Ultra 550B A55B | NVIDIA | 56.4 | 2 | 2026-06-17 |
| 58 | nvidia-nemotron-3-ultra-550b-a55b | NVIDIA | 56.4 | 1 | 2026-08-24 |
| 59 | Claude Sonnet 4.5 | Anthropic | 55.8 | 1 | 2026-07-02 |
| 60 | MiniMax 2.7 230B A10B | MiniMax | 55.5 | 1 | 2026-06-06 |
| 61 | MiniMax M2.7 | MiniMax | 55.4 | 3 | 2026-07-30 |
| 62 | Ling-3.0-flash | InclusionAI | 55.4 | 1 | 2026-08-06 |
| 63 | Gemini 3.5 Flash Lite | 54 | 4 | 2026-08-24 | |
| 64 | GPT-5.1 | OpenAI | 52.4 | 1 | 2026-07-02 |
| 65 | Grok Build 0.1 | SpaceXAI | 52.1 | 2 | 2026-07-07 |
| 66 | muse-glimmer-30b | Meta | 51.7 | 1 | 2026-08-23 |
| 67 | Muse Glimmer | Meta | 51.7 | 1 | 2026-08-11 |
| 68 | Qwen3.5 397B A17B | Alibaba | 51.3 | 2 | 2026-07-30 |
| 69 | LongCat-2.0 | Meituan | 50.2 | 1 | 2026-07-20 |
| 70 | Qwen 3.5 397B 17B | Alibaba | 49.9 | 1 | 2026-06-06 |
| 71 | GLM-4.6 | Z.ai | 49.4 | 1 | 2026-07-07 |
| 72 | Qwen3.5 122B A10B | Alibaba | 47.6 | 2 | 2026-07-09 |
| 73 | DeepSeek-V3.2 | DeepSeek | 46.8 | 1 | 2026-07-06 |
| 74 | Kimi K2.5 | Moonshot | 45.7 | 1 | 2026-07-14 |
| 75 | GLM-4.7 | Z.ai | 45.3 | 1 | 2026-07-02 |
| 76 | DeepSeek-V3.1-Terminus | DeepSeek | 44.9 | 1 | 2026-07-06 |
| 77 | Qwen3.6 35B A3B | Alibaba | 44.9 | 3 | 2026-08-11 |
| 78 | Claude Haiku 4.5 | Anthropic | 44.2 | 1 | 2026-06-17 |
| 79 | Solar Open2 250B | Upstage | 44.2 | 1 | 2026-08-13 |
| 80 | Gemma 4 31B | 43.4 | 2 | 2026-07-11 | |
| 81 | Qwen3.5 35B A3B | Alibaba | 40.8 | 1 | 2026-07-18 |
| 82 | K-EXAONE 2.0 0803 | LG AI | 40.4 | 1 | 2026-08-13 |
| 83 | Grok 4.3 | SpaceXAI | 39.7 | 2 | 2026-06-17 |
| 84 | Nemotron 3 Super | NVIDIA | 39.6 | 1 | 2026-08-11 |
| 85 | Step 3.7 Flash | StepFun | 39.3 | 2 | 2026-07-07 |
| 86 | Gemma 4 26B A4B | 39 | 2 | 2026-08-11 | |
| 87 | A.X-K2 | SK Telecom | 39 | 1 | 2026-08-13 |
| 88 | Nemotron 3 Super 120B A12B | NVIDIA | 38.6 | 1 | 2026-06-17 |
| 89 | Qwen3 Coder Next | Alibaba | 38.2 | 2 | 2026-07-07 |
| 90 | Claude Sonnet 4 | Anthropic | 36.3 | 1 | 2026-07-01 |
| 91 | North-Mini-Code-1.0 | Cohere | 35.6 | 2 | 2026-07-07 |
| 92 | GPT-5 | OpenAI | 35.2 | 1 | 2026-07-04 |
| 93 | GPT-5.5 Instant | OpenAI | 34.8 | 2 | 2026-07-07 |
| 94 | G9v3-39A5B | AI9Stars | 32.6 | 1 | 2026-08-04 |
| 95 | Gemini 3.1 Flash Lite Preview | 31.1 | 2 | 2026-07-07 | |
| 96 | Devstral 2 | Mistral | 30.3 | 2 | 2026-07-07 |
| 97 | K-EXAONE | LG AI | 30.3 | 1 | 2026-07-25 |
| 98 | Devstral Small 2 | Mistral | 29.6 | 2 | 2026-07-07 |
| 99 | Nova 2.0 Pro Preview | Amazon | 29.6 | 3 | 2026-06-24 |
| 100 | Qwen3.5 9B | Alibaba | 29.2 | 2 | 2026-06-17 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.