LMArena Agent Arena Code Confirmed Success
Agents · 2026-08-14
Signed percentage-point Agent Arena Code category diagnostic signal for sessions where the user explicitly confirmed task completion, recomputed on sessions tagged as writing or debugging code, automating workflows or analyzing data. 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 | 23.8 |
| Claude Opus 5.5 | 23.6 |
| Muse Wonder Full | 19.0 |
| GPT-6 Sol | 15.6 |
| GPT-6.1 Sol | 15.1 |
| Claude Fable 5.1 | 14.9 |
| GPT-6 Astra | 14.2 |
| Gemini 4 Argon | 12.9 |
| Claude Opus 5 | 12.7 |
| DeepSeek V4.1 Flash 1017 | 11.3 |
| Grok 4.7 | 8.3 |
| DeepSeek V4.1 Flash | 7.2 |
| Hy4 Preview | 6.8 |
| Muse Spark 1.3 | 6.7 |
| Muse Wonder | 6.2 |
| GLM-5.2 | 5.8 |
| MiMo-V2.6-Flash | 5.8 |
| MiMo-V2.6-Pro | 5.7 |
| Kimi K3 | 5.5 |
| Sonnet 5 | 5.3 |
| Qwen3.8-Flash-Next | 4.3 |
| GLM-5.3 | 3.9 |
| Claude Haiku 5.5 | 3.8 |
| Qwen3.8-Max | 3.6 |
| GPT-5.6 Sol | 3.4 |
| GLM-5.3 Flash | 2.7 |
| Claude Fable 5 | 2.6 |
| GPT-6 Luna | 2.3 |
| Mistral Large 4 Preview | 1.9 |
| Grok 4.6 | 0.3 |
| Step 5 Preview | 0.1 |
| Qwen3.8 27B | 0.0 |
| DeepSeek V4 Pro 0813 | -0.9 |
| GPT-5.5 | -1.0 |
| Gemini 3.8 Flash | -1.6 |
| GPT-5.4 | -3.8 |
| GPT-5.6 Luna | -5.4 |
| Gemini 3.7 Flash | -6.3 |
| Hy3 | -7.0 |
| MiMo-V2.5-Pro | -7.5 |