A benchmark for evaluating the agentic ability of models to complete practical tasks by using tools in a terminal environment.
Browse the latest scores, model modes, release dates, and parameter sizes for Terminal-Bench.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
| Rank | Model | License | |||
|---|---|---|---|---|---|
 GPT-5.1-Codex-Max Thinking Level · HighTools | 58.10 | 2025-11-19 | Unknown | Closed | |
GPT-5.1 Codex Thinking Level · HighTools | 56.30 | 2025-11-14 | Unknown | Closed | |
 Claude Sonnet 4.5 Thinking EnabledTools | 50.00 | 2025-09-30 | Unknown | Closed | |
4 |  Kimi K2 Thinking Thinking EnabledTools | 47.10 | 2025-11-06 | 1040B | Free Commercial |
5 | Opus 4.1 Extended Thinking | 46.50 | 2025-08-06 | Unknown | Closed |
6 | Kimi K2 0905 Standard Mode | 44.50 | 2025-09-05 | 1000B | Free Commercial |
7 | Kimi K2 0905 Thinking EnabledTools | 44.50 | 2025-09-05 | 1000B | Free Commercial |
8 | GPT-5 Thinking EnabledTools | 43.80 | 2025-08-07 | Unknown | Closed |
9 | Opus 4.1 Extended ThinkingTools | 43.30 | 2025-08-06 | Unknown | Closed |
10 | Claude Sonnet 4 Deep Thinking ModeTools | 41.30 | 2025-05-23 | Unknown | Closed |
11 | Haiku 4.5 32KTools | 41.00 | 2025-10-15 | Unknown | Closed |
12 |  GLM-4.6 Thinking EnabledTools | 40.50 | 2025-09-30 | 355B | Free Commercial |
13 |  Grok 4 Thinking EnabledTools | 38.00 | 2025-07-10 | Unknown | Closed |
14 |  DeepSeek V3.2-Exp Thinking EnabledTools | 37.70 | 2025-09-29 | 671B | Free Commercial |
15 | Kimi K2 Standard Mode | 37.50 | 2025-07-11 | 1000B | Free Commercial |
16 |  Qwen3-Coder-480B-A35B Standard Mode | 37.50 | 2025-07-23 | 480B | Free Commercial |
17 | GLM-4.5 Thinking Enabled | 37.50 | 2025-07-28 | 355B | Free Commercial |
18 | Claude Sonnet 4 Thinking EnabledTools | 35.50 | 2025-05-23 | Unknown | Closed |
19 | Qwen3-Coder-Flash Standard Mode | 31.30 | 2025-07-31 | 30.5B | Free Commercial |
20 | DeepSeek-V3.1 Standard ModeTools | 31.30 | 2025-08-20 | 671B | Free Commercial |
21 | OpenAI o3 Thinking Enabled | 30.20 | 2025-04-16 | Unknown | Closed |
22 | GLM-4.5-Air Thinking Enabled | 30.00 | 2025-07-28 | 106B | Free Commercial |
23 | DeepSeek-V3.1 Terminus Standard ModeTools | 30.00 | 2025-09-22 | 671B | Free Commercial |
24 | DeepSeek-V3.1 Terminus Thinking EnabledTools | 28.00 | 2025-09-22 | 671B | Free Commercial |
25 | Claude Sonnet 4.5 Standard ModeTools | 27.00 | 2025-09-30 | Unknown | Closed |
26 | Claude Sonnet 4 Standard ModeTools | 26.00 | 2025-05-23 | Unknown | Closed |
27 | Haiku 4.5 Standard ModeTools | 26.00 | 2025-10-15 | Unknown | Closed |
28 |  Gemini 2.5-Pro Thinking Enabled | 25.30 | 2025-06-05 | Unknown | Closed |
29 |  MiniMax M2 Thinking EnabledTools | 24.00 | 2025-10-27 | 230B | Free Commercial |
30 | DeepSeek V3.2-Exp Standard ModeTools | 23.00 | 2025-09-29 | 671B | Free Commercial |
