Digital Agent Task Execution on Claw-Anything
0.65Execution ScoreGPT-5.5
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-5.52026.05 | 0.65 | 34.5 | 53.5 | 20 | — | — | |
| Claude Opus 4.72026.05 | 0.62 | 31.8 | 48 | 13.5 | — | — | |
| Claw-Anything-Qwen3.5-27B# Params=27B2026.05 | 0.61 | 33.5 | 52 | 15.5 | — | — | |
| GLM-5.1# Params=754B2026.05 | 0.59 | 31.7 | 47 | 17 | — | — | |
| Claude Sonnet 4.52026.05 | 0.59 | 28 | 45 | 12 | — | — | |
| Qwen3.6-27B# Params=27B2026.05 | 0.58 | 22.5 | 42 | 6 | — | — | |
| Kimi-K2.6# Params=1.1T2026.05 | 0.57 | 22.8 | 44 | 6.5 | — | — | |
| MiniMax-M2.7# Params=229B2026.05 | 0.52 | 13.5 | 28.5 | 3.5 | — | — | |
| Qwen3.5-27B# Params=27B2026.05 | 0.5 | 9.8 | 19 | 2 | — | — |