Terminal Bench 2.0
本文介绍 Terminal-Bench 的设计理念，深入讲解 core、Terminal-Bench Hard 与最新 Terminal-Bench 2.0 的区别，帮助开发者选择合适的 AI 终端评测基准。
更新于 2026-04-08
8 次浏览
问题数量
89
发布机构
个人
评测类别
AI Agent - 工具使用
评测指标
Accuracy
支持语言
英文
难度等级
高难度
简介
用于评估大模型在终端环境下使用工具能力的评测基准的升级版本
相关资源
Terminal Bench 2.0 Model Score Leaderboard
Source: DataLearnerAI
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
模式说明:
normal
thinking
low
medium
high
deeper thinking
parallel_thinking
图表加载中...
Latest Terminal Bench 2.0 model rankings and full benchmark leaderboard
Browse the latest scores, model modes, release dates, and parameter sizes for Terminal Bench 2.0.
Terminal Bench 2.0详细排名数据表格
| 排名 | 模型 | |||
|---|---|---|---|---|
| 1 | Claude Mythos Preview Deep ThinkingTools | 82 | 2026-04-07 | 未知 |
| 2 | GPT-5.3 Codex Deep ThinkingTools | 77.3 | 2026-02-05 | 未知 |
| 3 | GPT-5.4 Deep ThinkingTools | 75.1 | 2026-03-05 | 未知 |
| 4 | Gemini 3.1 Pro Preview Thinking Level · HighTools | 68.5 | 2026-02-20 | 未知 |
| 5 | Claude Opus 4.6 Deep ThinkingTools | 65.4 | 2026-02-05 | 未知 |
| 6 | GLM 5.1 Thinking Level · MediumTools | 63.5 | 2026-03-27 | 754 |
| 7 | Composer 2 Thinking Level · Medium | 61.7 | 2026-03-19 | 未知 |
| 8 | Qwen 3.6 Plus Preview Thinking Level · MediumTools | 61.6 | 2026-03-31 | 未知 |
| 9 | GLM-5 Thinking Level · MediumTools | 61.1 | 2026-02-11 | 7440 |
| 10 | GPT-5.4 mini Deep ThinkingTools | 60 | 2026-03-17 | 未知 |
| 11 | Claude Opus 4.5 Thinking Level · MediumTools | 59.3 | 2025-11-25 | 未知 |
| 12 | Claude Sonnet 4.6 Thinking Level · MediumTools | 59.1 | 2026-02-17 | 未知 |
| 13 | Gemini 3.0 Pro (Preview 11-2025) Thinking Level · HighTools | 56.9 | 2025-11-18 | 未知 |
| 14 | Gemini 3.0 Pro (Preview 11-2025) Thinking Level · MediumTools | 54.2 | 2025-11-18 | 未知 |
| 15 | Qwen3.5-397B-A17B Thinking Level · MediumTools | 52.5 | 2026-02-16 | 397 |
| 16 | MiniMax M2.5 Thinking Level · MediumTools | 51.7 | 2026-02-12 | 2290 |
| 17 | Step 3.5 Flash Thinking Level · MediumTools | 51 | 2026-02-02 | 1960 |
| 18 | Kimi K2.5 Thinking Level · MediumTools | 50.8 | 2026-01-27 | 10000 |
| 19 | M2.1 Thinking Level · MediumTools | 47.9 | 2025-12-23 | 2300 |
| 20 | Composer 1.5 Thinking Level · Medium | 47.9 | 2026-02-09 | 未知 |
| 21 | Gemini 3.0 Flash Thinking Level · MediumTools | 47.6 | 2025-12-17 | 未知 |
| 22 | GPT-5.1 Thinking Level · HighTools | 47.6 | 2025-11-12 | 未知 |
| 23 | DeepSeek V3.2 Thinking Level · MediumTools | 46.4 | 2025-12-01 | 6710 |
| 24 | GPT-5.4 nano Deep ThinkingTools | 46.3 | 2026-03-17 | 未知 |
| 25 | Claude Sonnet 4.5 Thinking Level · MediumTools | 42.8 | 2025-09-30 | 未知 |
| 26 | Qwen3.5-27B Thinking Level · MediumTools | 41.6 | 2026-02-25 | 270 |
| 27 | GLM-4.7 Thinking Level · MediumTools | 41 | 2025-12-22 | 3580 |
| 28 | Composer 1 Thinking Level · Medium | 40 | 2025-10-29 | 未知 |
| 29 | Qwen3-Coder-Next NormalTools | 36.2 | 2026-02-03 | 80 |
| 30 | Gemini 2.5-Pro Thinking Level · MediumTools | 32.6 | 2025-06-05 | 未知 |