LMArena Agent Arena Code Praise vs Complaint
Agents · 2026-08-14
Signed percentage-point Agent Arena Code category diagnostic signal comparing positive user feedback against complaints, 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 Opus 5.5 | 49.7 |
| GPT-6 Astra | 31.4 |
| GPT-6.1 Sol | 29.9 |
| Claude Sonnet 5.5 | 24.5 |
| Claude Fable 5.1 | 23.5 |
| Gemini 4 Argon | 21.1 |
| Kimi K3.1 Pro | 19.0 |
| Muse Wonder Full | 17.6 |
| Claude Fable 5 | 14.5 |
| GPT-6 Sol | 13.8 |
| Claude Opus 5 | 8.8 |
| DeepSeek V4.1 Flash 1017 | 8.8 |
| GPT-5.6 Sol | 8.7 |
| GLM-5.2 | 6.2 |
| Sonnet 5 | 5.4 |
| GLM-5.3 | 4.4 |
| GPT-5.6 Terra | 4.0 |
| Grok 4.7 | 3.4 |
| MiMo-V2.6-Flash | 3.3 |
| DeepSeek V4.1 Flash | 3.0 |
| Gemini 3.8 Flash | 2.7 |
| Muse Spark 1.3 | 2.6 |
| MiMo-V2.6-Pro | 2.3 |
| Hy4 Preview | 2.2 |
| GPT-6 Luna | 2.1 |
| Claude Haiku 5.5 | 1.9 |
| Qwen3.8-Max | 1.1 |
| Muse Wonder | 0.8 |
| Kimi K3 | 0.1 |
| Grok 4.6 | -1.2 |
| DeepSeek V4 Pro 0813 | -1.2 |
| Step 5 Preview | -1.3 |
| GLM-5.3 Flash | -2.9 |
| Qwen3.8-Flash-Next | -3.0 |
| Mistral Large 4 Preview | -3.2 |
| Gemini 3.7 Flash | -3.5 |
| Hy3 | -4.6 |
| Qwen3.8 27B | -5.5 |
| Gemini 3.1 Pro Preview | -6.2 |
| Qwen3.7-Max (source unspecified) | -10.3 |