WeaveBench — leaderboard
Metric: PassRate (%). Source: weavebench.github.io. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 35.1 |
| 2 | GPT-5.5 | 33.3 |
| 3 | GPT-5.4 | 22.8 |
| 4 | GPT-5.3 Codex | 18.4 |
| 5 | GPT-5.2 Codex | 6.1 |
| 6 | Gemini 3.1 Pro (Preview) | 1.8 |
| 7 | GPT-5.1 Codex | 1.8 |
| 8 | Qwen 3.5 397B A17B | 0.9 |
| 9 | Qwen 3 VL 8B (Thinking) | 0.9 |
Interactive version: aibenchmarks.dev/benchmark?slug=weavebench · How the rankings work · Data refreshed daily, snapshot 2026-07-20.