Computer Anthology Terminal Tasks (Terminus-2): leaderboard
Metric: pass@1 (%). Source: vetto.ai. 28 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6 (Medium) | 67 |
| 2 | Claude Fable 5.1 (High) | 62.4 |
| 3 | Claude Opus 5 (High) | 62 |
| 4 | Gemini 3.8 Flash (Medium) | 51.8 |
| 5 | GPT-5.6 Sol (Medium) | 51.6 |
| 6 | Claude Fable 5 (High) | 50.8 |
| 7 | Grok 4.6 (High) | 49.2 |
| 8 | Gemini 3.7 Flash (Medium) | 46 |
| 9 | Kimi K3 (Max) | 40 |
| 10 | Muse Spark 1.1 | 39 |
| 11 | GPT-5.5 (Medium) | 38.6 |
| 12 | DeepSeek V4 Pro (0813) (High) | 32.6 |
| 13 | GLM-5.2 (High) | 31.4 |
| 14 | Claude Opus 4.8 (High) | 29.8 |
| 15 | Qwen 3.8 Max (xHigh) | 26 |
Interactive version: theaggregate.ai/benchmark?slug=computer-anthology-terminal-tasks-terminus-2 · How It Works · Data refreshed daily, snapshot 2026-09-20.