TestsBenchmarks · Agentic
Terminal-Bench 2.0
Agentic tasks in a command-line environment (planning, iteration, tool use).
Measured by independent testersIndependent Reported by the makerVendor-reported
| Thinking levelSetting | |||||||
|---|---|---|---|---|---|---|---|
| 1 | GPT-5.5 | OpenAI | 82.7% | Extra highreasoning effort=xhigh | Maker's own figureVendor-reportedOpenAI ↗ | Codex | 23 Apr 2026 |
| 2 | GPT-5.4 | OpenAI | 75.1% | Extra highreasoning effort=xhigh | Maker's own figureVendor-reportedOpenAI ↗ | — | 23 Apr 2026 |
| 3 | Qwen3.7 Plus | Qwen (Alibaba) | 70.3% | Defaultthinking | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 (Harbor) | 21 May 2026 |
| 4 | Qwen3.7 Max | Qwen (Alibaba) | 69.7% | Defaultthinking (vendor recommends 'Reasoning effort is set to xhigh' system prompt for reasoning tasks) | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 (Harbor) | 16 May 2026 |
| 5 | Gemini 3.1 Pro (Preview) | Google (Gemini / DeepMind) | 68.5% | HighThinking (High) | Maker's own figureVendor-reportedGoogle ↗ | Terminus-2 | 19 Feb 2026 |
| 6 | DeepSeek V4 Pro (Preview, 0423) | DeepSeek | 67.9% | MaxThink Max | Maker's own figureVendor-reportedDeepSeek ↗ | — | 24 Apr 2026 |
| 7 | Kimi K2.6 | Moonshot AI (Kimi) | 66.7% | Defaultthinking | Maker's own figureVendor-reportedMoonshot AI ↗ | Terminus 2 | 20 Apr 2026 |
| 8 | GLM-5.1 | Z.ai (Zhipu) | 63.5% | Defaultthinking | Maker's own figureVendor-reportedZ.ai ↗ | Terminus 2 | 7 Apr 2026 |
| 9 | Qwen3.6 Plus | Qwen (Alibaba) | 61.6% | Defaultthinking | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 (Harbor) | 2 Apr 2026 |
| 10 | Qwen3.6 27B | Qwen (Alibaba) | 59.3% | Defaultthinking | Maker's own figureVendor-reportedQwen (Alibaba) ↗ | Terminus 2 (Harbor) | 22 Apr 2026 |
| 11 | MiniMax M2.7 | MiniMax | 57.0% | Defaultsetting not stated | Maker's own figureVendor-reportedMiniMax ↗ | — | 18 Mar 2026 |
| 12 | DeepSeek V4 Flash (Preview, 0423) | DeepSeek | 56.9% | MaxThink Max | Maker's own figureVendor-reportedDeepSeek ↗ | — | 24 Apr 2026 |