LoopsBench - Claude Code Loop - Test Pass Rate (Single Segment): leaderboard
Metric: Released-test pass rate (%; without outer-loop continuation). Source: arxiv.org. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 38.04 |
| 2 | DeepSeek V4 Pro | 36.07 |
| 3 | GLM-5.1 | 34.55 |
| 4 | Gemini 3.1 Pro (Preview) | 32.86 |
| 5 | Grok 4.1 Fast (Reasoning) | 23.32 |
| 6 | Kimi K2.6 | 21.27 |
Interactive version: theaggregate.ai/benchmark?slug=loopsbench-claude-code-loop-test-pass-rate-single-segment · How It Works · Data refreshed daily, snapshot 2026-09-19.