# Terminal-Bench 2.0 leaderboard

Coding benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/coding-terminal-bench2

Every model the catalog carries a published Terminal-Bench 2.0 value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Coding |
| Measure | Interactive CLI agent evaluation |
| Tasks | Terminal-based software tasks |
| Difficulty | Professional software engineering |
| Published by | [Terminal-Bench 2.0](https://www.tbench.ai/) |

A benchmark for agentic software engineering tasks executed in real terminal environments. DeepSeek reports it in the agentic section, while BenchLM also mirrors it in coding for models that publish it as a developer-task signal.

## Terminal-Bench 2.0 ranking

66 models with a published Terminal-Bench 2.0 value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Interactive CLI agent evaluation | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [GPT-5.6 Sol](https://modelscale.dev/models/gpt-5-6-sol) | OpenAI | 91.9 | Last good 2026-09-23 · source benchlm:models |
| 2 | [Claude Mythos 5](https://modelscale.dev/models/claude-mythos-5) | Anthropic | 88.0 | Last good 2026-09-23 · source benchlm:models |
| 3 | [GPT-5.6 Terra](https://modelscale.dev/models/gpt-5-6-terra) | OpenAI | 87.4 | Last good 2026-09-23 · source benchlm:models |
| 4 | [GPT-5.6 Luna](https://modelscale.dev/models/gpt-5-6-luna) | OpenAI | 84.7 | Last good 2026-09-23 · source benchlm:models |
| 5 | [Claude Fable 5](https://modelscale.dev/models/claude-fable-5) | Anthropic | 84.3 | Last good 2026-09-23 · source benchlm:models |
| 6 | [Grok 4.5](https://modelscale.dev/models/grok-4-5) | xAI | 83.3 | Last good 2026-09-23 · source benchlm:models |
| 7 | [Sakana Fugu-Ultra](https://modelscale.dev/models/sakana-fugu-ultra) | Sakana AI | 82.1 | Last good 2026-09-23 · source benchlm:models |
| 8 | [GPT-5.5](https://modelscale.dev/models/gpt-5-5) | OpenAI | 82.0 | Observed 2026-11-21 · source benchlm:models |
| 9 | [SWE-1.7](https://modelscale.dev/models/swe-1-7) | Cognition | 81.5 | Last good 2026-09-23 · source benchlm:models |
| 10 | [GLM-5.2](https://modelscale.dev/models/glm-5-2) | Z.AI | 81.0 | Last good 2026-09-23 · source benchlm:models |
| 11 | [Claude Sonnet 5](https://modelscale.dev/models/claude-sonnet-5) | Anthropic | 80.4 | Last good 2026-09-23 · source benchlm:models |
| 12 | [Sakana Fugu](https://modelscale.dev/models/sakana-fugu) | Sakana AI | 80.2 | Last good 2026-09-23 · source benchlm:models |
| 13 | [Muse Spark 1.1](https://modelscale.dev/models/muse-spark-1-1) | Meta | 80.0 | Last good 2026-09-23 · source benchlm:models |
| 14 | [Ornith-1.0-397B](https://modelscale.dev/models/ornith-1-0-397b) | DeepReinforce AI | 77.5 | Last good 2026-09-23 · source benchlm:models |
| 15 | [GPT-5.3 Codex](https://modelscale.dev/models/gpt-5-3-codex) | OpenAI | 77.3 | Observed 2026-11-21 · source benchlm:models |
| 16 | [Gemini 3.5 Flash](https://modelscale.dev/models/gemini-3-5-flash) | Google | 76.2 | Last good 2026-09-23 · source benchlm:models |
| 17 | [GPT-5.4](https://modelscale.dev/models/gpt-5-4) | OpenAI | 75.1 | Observed 2026-11-21 · source benchlm:models |
| 18 | [Claude Opus 4.8](https://modelscale.dev/models/claude-opus-4-8) | Anthropic | 74.6 | Last good 2026-09-23 · source benchlm:models |
| 19 | [Qwen3.7 Plus](https://modelscale.dev/models/qwen3-7-plus) | Alibaba | 70.3 | Observed 2026-11-21 · source benchlm:models |
| 20 | [Laguna S 2.1](https://modelscale.dev/models/laguna-s-2-1) | Poolside | 70.2 | Last good 2026-09-23 · source benchlm:models |
| 21 | [Qwen3.7 Max](https://modelscale.dev/models/qwen3-7-max) | Alibaba | 69.7 | Observed 2026-11-21 · source benchlm:models |
| 22 | [Claude Opus 4.7](https://modelscale.dev/models/claude-opus-4-7) | Anthropic | 69.4 | Observed 2026-11-21 · source benchlm:models |
| 22 | [Claude Opus 4.7 (Adaptive)](https://modelscale.dev/models/claude-opus-4-7-max) | Anthropic | 69.4 | Last good 2026-11-18 · source benchlm:models |
| 24 | [Composer 2.5](https://modelscale.dev/models/composer-2-5) | Cursor | 69.3 | Observed 2026-11-21 · source benchlm:models |
| 25 | [MiMo-V2.5-Pro](https://modelscale.dev/models/mimo-v2-5-pro) | Xiaomi | 68.4 | Observed 2026-11-21 · source benchlm:models |
| 26 | [DeepSeek V4 Pro](https://modelscale.dev/models/deepseek-v4-pro) | DeepSeek | 67.9 | Observed 2026-11-21 · source benchlm:models |
| 26 | [DeepSeek V4 Pro 0813](https://modelscale.dev/models/deepseek-v4-pro-max) | DeepSeek | 67.9 | Last good 2026-11-18 · source benchlm:models |
| 28 | [Kimi K2.6](https://modelscale.dev/models/kimi-2-6) | Moonshot AI | 66.7 | Observed 2026-11-21 · source benchlm:models |
| 29 | [MiniMax M3](https://modelscale.dev/models/minimax-m3) | MiniMax | 66.0 | Last good 2026-09-23 · source benchlm:models |
| 30 | [MiMo-V2.5](https://modelscale.dev/models/mimo-v2-5) | Xiaomi | 65.8 | Observed 2026-11-21 · source benchlm:models |
| 31 | [Claude Opus 4.6](https://modelscale.dev/models/claude-opus-4-6) | Anthropic | 65.4 | Observed 2026-11-21 · source benchlm:models |
| 31 | [Qwen 3.6 Max (preview)](https://modelscale.dev/models/qwen3-6-max-preview) | Alibaba | 65.4 | Observed 2026-11-21 · source benchlm:models |
| 33 | [Inkling-Small](https://modelscale.dev/models/inkling-small) | Thinking Machines Lab | 64.7 | Last good 2026-09-23 · source benchlm:models |
| 34 | [Ornith-1.0-35B](https://modelscale.dev/models/ornith-1-0-35b) | DeepReinforce AI | 64.2 | Last good 2026-09-23 · source benchlm:models |
| 35 | [Inkling](https://modelscale.dev/models/inkling) | Thinking Machines Lab | 63.8 | Last good 2026-09-23 · source benchlm:models |
| 36 | [GLM-5.1](https://modelscale.dev/models/glm-5-1) | Z.AI | 63.5 | Observed 2026-11-21 · source benchlm:models |
| 37 | [Composer 2](https://modelscale.dev/models/composer-2) | Cursor | 61.7 | Observed 2026-11-21 · source benchlm:models |
| 38 | [Qwen3.6 Plus](https://modelscale.dev/models/qwen3-6-plus) | Alibaba | 61.6 | Observed 2026-11-21 · source benchlm:models |
| 39 | [GPT-5.4 mini](https://modelscale.dev/models/gpt-5-4-mini) | OpenAI | 60.0 | Observed 2026-11-21 · source benchlm:models |
| 40 | [Step 3.7 Flash](https://modelscale.dev/models/step-3-7-flash) | StepFun | 59.5 | Last good 2026-09-23 · source benchlm:models |
| 41 | [Claude Opus 4.5](https://modelscale.dev/models/claude-opus-4-5) | Anthropic | 59.3 | Observed 2026-11-21 · source benchlm:models |
| 41 | [Qwen3.6-27B](https://modelscale.dev/models/qwen3-6-27b) | Alibaba | 59.3 | Observed 2026-11-21 · source benchlm:models |
| 43 | [Claude Sonnet 4.6](https://modelscale.dev/models/claude-sonnet-4-6) | Anthropic | 59.1 | Observed 2026-11-21 · source benchlm:models |
| 44 | [MiniMax M2.7](https://modelscale.dev/models/minimax-m2-7) | MiniMax | 57.0 | Observed 2026-11-21 · source benchlm:models |
| 45 | [DeepSeek V4 Flash 0731](https://modelscale.dev/models/deepseek-v4-flash-max) | DeepSeek | 56.9 | Observed 2026-11-21 · source benchlm:models |
| 46 | [Nemotron 3 Ultra](https://modelscale.dev/models/nemotron-3-ultra-500b) | NVIDIA | 56.4 | Last good 2026-09-23 · source benchlm:models |
| 47 | [GLM-5](https://modelscale.dev/models/glm-5) | Z.AI | 56.2 | Observed 2026-11-21 · source benchlm:models |
| 48 | [Hy3 Preview](https://modelscale.dev/models/hy3-preview) | Tencent | 54.4 | Observed 2026-11-21 · source benchlm:models |
| 49 | [Gemini 3.5 Flash-Lite](https://modelscale.dev/models/gemini-3-5-flash-lite) | Google | 54.0 | Last good 2026-09-23 · source benchlm:models |
| 50 | [Qwen3.5 397B](https://modelscale.dev/models/qwen3-5-397b) | Alibaba | 52.5 | Observed 2026-11-21 · source benchlm:models |
| 51 | [Qwen3.6-35B-A3B](https://modelscale.dev/models/qwen3-6-35b-a3b) | Alibaba | 51.5 | Observed 2026-11-21 · source benchlm:models |
| 52 | [Kimi K2.5](https://modelscale.dev/models/kimi-k2-5) | Moonshot AI | 50.8 | Observed 2026-11-21 · source benchlm:models |
| 52 | [Kimi K2.5 (Reasoning)](https://modelscale.dev/models/kimi-k2-5-reasoning) | Moonshot AI | 50.8 | Last good 2026-11-18 · source benchlm:models |
| 54 | [Claude Sonnet 4.5](https://modelscale.dev/models/claude-sonnet-4-5) | Anthropic | 50.0 | Observed 2026-11-21 · source benchlm:models |
| 55 | [Qwen3.5-122B-A10B](https://modelscale.dev/models/qwen3-5-122b-a10b) | Alibaba | 49.4 | Observed 2026-11-21 · source benchlm:models |
| 56 | [GPT-5.4 nano](https://modelscale.dev/models/gpt-5-4-nano) | OpenAI | 46.3 | Observed 2026-11-21 · source benchlm:models |
| 57 | [MAI-Thinking-1](https://modelscale.dev/models/mai-thinking-1) | Microsoft | 46.0 | Observed 2026-11-21 · source benchlm:models |
| 58 | [Laguna M.1](https://modelscale.dev/models/laguna-m-1) | Poolside | 45.8 | Observed 2026-11-21 · source benchlm:models |
| 59 | [Ornith-1.0-9B](https://modelscale.dev/models/ornith-1-0-9b) | DeepReinforce AI | 43.1 | Last good 2026-09-23 · source benchlm:models |
| 60 | [Qwen3.5-27B](https://modelscale.dev/models/qwen3-5-27b) | Alibaba | 41.6 | Observed 2026-11-21 · source benchlm:models |
| 61 | [GLM-4.7](https://modelscale.dev/models/glm-4-7) | Z.AI | 41.0 | Observed 2026-11-21 · source benchlm:models |
| 62 | [Qwen3.5-35B-A3B](https://modelscale.dev/models/qwen3-5-35b-a3b) | Alibaba | 40.5 | Observed 2026-11-21 · source benchlm:models |
| 63 | [Laguna XS 2.1](https://modelscale.dev/models/laguna-xs-2-1) | Poolside | 37.5 | Observed 2026-11-21 · source benchlm:models |
| 64 | [Laguna XS.2](https://modelscale.dev/models/laguna-xs-2) | Poolside | 35.7 | Observed 2026-11-21 · source benchlm:models |
| 65 | [LongCat-Flash-Lite-Sparse](https://modelscale.dev/models/longcat-flash-lite-sparse) | Meituan | 33.7 | Observed 2026-11-21 · source benchlm:models |
| 66 | [Nemotron 3.5 Lightning 30B A3B NVFP4](https://modelscale.dev/models/nemotron-3-5-lightning-30b-a3b-nvfp4) | NVIDIA | 23.5 | Last good 2026-09-23 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
