Finding the Right Fit - Terminal-Bench 4 (63 Tasks, OpenHands): leaderboard
A 63-task subset of Terminal-Bench 4 (the tasks that need no H100 GPU), run by the Finding the Right Fit harness study with every model inside the OpenHands agent harness at high reasoning effort, one counted run per task. Percent of tasks solved, higher is better (Li et al., 2026).
Metric: Task score (%), OpenHands harness, high reasoning effort, one run per task over 63 tasks. Source: huggingface.co. Saturation forecast: Around February 2027. 5 models tracked. Updated 5 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 5 | High | 57.14 | #26 |
| 2 | GPT-6 Astra | High | 49.21 | #10 |
| 3 | GLM-5.3 | High | 41.27 | #73 |
| 4 | Kimi K3 | High | 11.11 | #61 |
| 5 | DeepSeek V4 Pro | High | 3.17 | #198 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=finding-the-right-fit-terminal-bench-4-63-tasks-openhands · How It Works · Data refreshed daily, snapshot 2026-11-20.