WildClawBench: leaderboard
60 hand-built agent tasks run inside the OpenClaw assistant with real shells, files, browsers, email and calendar, spanning productivity, coding, social, search, creative and safety; InternLM, 2026.
Metric: Overall Score (verified). Source: internlm.github.io. Source: benchmarklist.com. Saturation forecast: Around July 2027. 72 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 67.2 | #19 |
| 2 | Claude Opus 4.8 | 64.7 | #37 |
| 3 | Seed 2.1 Turbo | 62.8 | #86 |
| 4 | Claude Opus 4.7 | 62.2 | #49 |
| 5 | Claude Fable 5 | 62 | #16 |
| 6 | GPT-5.5 | 58.2 | #31 |
| 7 | Grok 4.5 | 57.5 | #48 |
| 8 | Qwen 3.8 Max | 56.2 | #35 |
| 9 | Muse Spark 1.1 | 54.8 | #52 |
| 10 | Kimi K3 | 54.5 | #30 |
| 11 | GLM-5.2 | 54.2 | #73 |
| 12 | Nex N2 Pro | 53.5 | #107 |
| 13 | Claude Opus 4.6 | 51.6 | #63 |
| 14 | GPT-5.4 | 50.3 | #79 |
| 15 | Hy3 | 49.7 | #93 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=wildclawbench · How It Works · Data refreshed daily, snapshot 2026-10-20.