LMArena Agent Arena Work Confirmed Success
Agents · 2026-08-14
Signed percentage-point Agent Arena Work category diagnostic signal for sessions where the user explicitly confirmed task completion, recomputed on sessions tagged as office work such as document creation, professional research, planning and professional writing. Values may be negative relative to the category baseline; reported confidence intervals are not standard errors.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Sonnet 5.5 | 22.1 |
| Claude Opus 5.5 | 19.0 |
| GPT-6.1 Sol | 13.5 |
| Claude Fable 5.1 | 12.4 |
| Gemini 4 Argon | 11.4 |
| GPT-6 Astra | 11.2 |
| Muse Wonder Full | 10.9 |
| Claude Opus 5 | 10.2 |
| Muse Wonder | 8.8 |
| DeepSeek V4.1 Flash 1017 | 7.5 |
| MiMo-V2.6-Pro | 6.1 |
| GPT-6 Sol | 6.0 |
| MiMo-V2.6-Flash | 5.9 |
| DeepSeek V4.1 Flash | 5.7 |
| Kimi K3 | 5.6 |
| GLM-5.2 | 5.5 |
| Hy4 Preview | 5.5 |
| Muse Spark 1.3 | 5.0 |
| Grok 4.7 | 3.8 |
| GLM-5.3 | 3.5 |
| Qwen3.8-Max | 2.9 |
| GPT-5.6 Sol | 2.8 |
| Kimi K3.1 Pro | 2.8 |
| Claude Fable 5 | 2.2 |
| GLM-5.3 Flash | 2.2 |
| Sonnet 5 | 2.2 |
| Claude Haiku 5.5 | 2.2 |
| Qwen3.8-Flash-Next | 1.7 |
| GPT-6 Luna | 1.0 |
| DeepSeek V4 Pro 0813 | 0.9 |
| Step 5 Preview | 0.6 |
| Gemini 3.8 Flash | -0.3 |
| Mistral Large 4 Preview | -1.6 |
| Qwen3.8 27B | -1.9 |
| Grok 4.6 | -2.4 |
| GPT-5.6 Terra | -3.4 |
| Gemini 3.7 Flash | -5.8 |
| Muse Spark 1.1 | -7.9 |
| Qwen3.7-Max (source unspecified) | -8.3 |
| MiMo-V2.5-Pro | -8.4 |