# Terminal-Bench 4.0 leaderboard

Agentic 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/agentic-terminal-bench4

Every model the catalog carries a published Terminal-Bench 4.0 value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Agentic |
| Measure | Task completion rate across 5 trials per task |
| Tasks | 66 professional computer-work tasks |
| Difficulty | Frontier autonomous knowledge work |
| Published by | [Terminal-Bench 4.0](https://www.tbench.ai/news/terminal-bench-4-0) |

Terminal-Bench 4.0 measures difficult computer work after recalibrating task resources, fixing unstable tasks, and removing tasks that no longer separate frontier systems. Version 4.1 regrades the same trials with hardened verifiers.

## Terminal-Bench 4.0 ranking

29 models with a published Terminal-Bench 4.0 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 | Task completion rate across 5 trials per task | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Claude Sonnet 5.5](https://modelscale.dev/models/claude-sonnet-5-5) | Anthropic | 70.60 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Claude Mythos 5.2](https://modelscale.dev/models/claude-mythos-5-2) | Anthropic | 70.50 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Claude Fable 5.2](https://modelscale.dev/models/claude-fable-5-2) | Anthropic | 67.60 | Observed 2026-11-21 · source benchlm:models |
| 4 | [Claude Opus 5.5](https://modelscale.dev/models/claude-opus-5-5) | Anthropic | 66.40 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Kimi K3.1 Pro](https://modelscale.dev/models/kimi-3-1-pro) | Moonshot AI | 65.70 | Observed 2026-11-21 · source benchlm:models |
| 6 | [Claude Mythos 5.1](https://modelscale.dev/models/claude-mythos-5-1) | Anthropic | 60.90 | Observed 2026-11-21 · source benchlm:models |
| 7 | [GPT-6 Astra](https://modelscale.dev/models/gpt-6-astra) | OpenAI | 57.90 | Observed 2026-11-21 · source benchlm:models |
| 8 | [Gemini 4 Argon](https://modelscale.dev/models/gemini-4-argon) | Google | 57.40 | Observed 2026-11-21 · source benchlm:models |
| 9 | [Claude Fable 5.1](https://modelscale.dev/models/claude-fable-5-1) | Anthropic | 55.80 | Observed 2026-11-21 · source benchlm:models |
| 10 | [Pareto 26.9](https://modelscale.dev/models/pareto-26-9) | Unbiased | 51.00 | Observed 2026-11-21 · source benchlm:models |
| 11 | [Pareto 26.10 Preview](https://modelscale.dev/models/pareto-26-10-preview) | Unbiased | 50.80 | Observed 2026-11-21 · source benchlm:models |
| 12 | [GPT-6.1 Terra](https://modelscale.dev/models/gpt-6-1-terra) | OpenAI | 50.30 | Observed 2026-11-21 · source benchlm:models |
| 13 | [Qwen4 Max](https://modelscale.dev/models/qwen4-max) | Alibaba | 48.50 | Observed 2026-11-21 · source benchlm:models |
| 14 | [ERNIE 5.5 Flash Preview](https://modelscale.dev/models/ernie-5-5-flash-preview) | Baidu | 46.50 | Observed 2026-11-21 · source benchlm:models |
| 15 | [Hy4](https://modelscale.dev/models/hy4) | Tencent | 44.40 | Observed 2026-11-21 · source benchlm:models |
| 16 | [DeepSeek V4.1 Flash 1017](https://modelscale.dev/models/deepseek-v4-1-flash) | DeepSeek | 42.30 | Observed 2026-11-21 · source benchlm:models |
| 17 | [Muse Ember](https://modelscale.dev/models/muse-ember) | Meta | 40.90 | Observed 2026-11-21 · source benchlm:models |
| 18 | [Claude Haiku 5.5](https://modelscale.dev/models/claude-haiku-5-5) | Anthropic | 39.20 | Observed 2026-11-21 · source benchlm:models |
| 19 | [LongCat-2.5](https://modelscale.dev/models/longcat-2-5) | Meituan | 38.50 | Observed 2026-11-21 · source benchlm:models |
| 20 | [MiniMax M3.1 Flash](https://modelscale.dev/models/minimax-m3-1-flash) | MiniMax | 38.10 | Observed 2026-11-21 · source benchlm:models |
| 21 | [Grok 4.7](https://modelscale.dev/models/grok-4-7) | xAI | 38.00 | Observed 2026-11-21 · source benchlm:models |
| 22 | [MiMo-V2.6-Pro](https://modelscale.dev/models/mimo-v2-6-pro) | Xiaomi | 34.90 | Observed 2026-11-21 · source benchlm:models |
| 23 | [Step 5 Preview](https://modelscale.dev/models/step-5-preview) | StepFun | 33.30 | Observed 2026-11-21 · source benchlm:models |
| 24 | [MiMo-V2.6-Flash](https://modelscale.dev/models/mimo-v2-6-flash) | Xiaomi | 28.80 | Observed 2026-11-21 · source benchlm:models |
| 25 | [SWE-2](https://modelscale.dev/models/swe-2) | Cognition | 27.30 | Observed 2026-11-21 · source benchlm:models |
| 26 | [SWE-2 Flash](https://modelscale.dev/models/swe-2-flash) | Cognition | 20.00 | Observed 2026-11-21 · source benchlm:models |
| 27 | [Hy4 flash](https://modelscale.dev/models/hy4-flash) | Tencent | 19.20 | Observed 2026-11-21 · source benchlm:models |
| 28 | [Gemini 3.8 Flash](https://modelscale.dev/models/gemini-3-8-flash) | Google | 19.10 | Observed 2026-11-21 · source benchlm:models |
| 29 | [Inkling 2 Mini](https://modelscale.dev/models/inkling-2-mini) | Thinking Machines Lab | 5.60 | Observed 2026-11-21 · 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.
