| # | Model | Arena score | Samples |
|---|---|---|---|
| 1 |  claude-fable-5Anthropic | 1,506±5 | 29,683 |
| 2 | claude-opus-4-6-highAnthropic | 1,505±4 | 71,992 |
| 3 | claude-opus-4-7-highAnthropic | 1,502±4 | 59,992 |
| 4 | claude-fable-5.1-maxAnthropic | 1,501±8 | 5,447 |
| 5 | 1,499±11 | 3,229 | |
| 6 | claude-opus-4-6Anthropic | 1,497±3 | 75,898 |
| 7 |
| claude-opus-4-7Anthropic |
|---|
| 1,494±4 |
| 61,138 |
| 8 | muse-spark-1.1Meta | 1,494±5 | 27,136 |
|---|
| 9 | claude-opus-5-highAnthropic | 1,493±4 | 41,777 |
|---|
| 10 |  gemini-3.8-flash-highGoogleEarly | 1,493±9 | 5,094 |
|---|
| 11 | gemini-3.7-flash-highGoogleEarly | 1,490±8 | 5,645 |
|---|
| 12 | muse-sparkMeta | 1,488±6 | 13,565 |
|---|
| 13 | gemini-3.1-pro-previewGoogle | 1,487±3 | 106,483 |
|---|
| 14 | claude-opus-5-maxAnthropic | 1,487±6 | 20,296 |
|---|
| 15 |  glm-5.3-maxZ.ai | 1,486±7 | 10,559 |
|---|
| 16 | gemini-3-proGoogle | 1,486±4 | 40,653 |
|---|
| 17 |  kimi-k3-maxMoonshot | 1,485±5 | 20,911 |
|---|
| 18 | claude-opus-4-8-highAnthropic | 1,482±4 | 52,123 |
|---|
| 19 |  gpt-5.6-sol-xhighOpenAI | 1,482±5 | 26,611 |
|---|
| 20 | gpt-5.5-highOpenAI | 1,482±4 | 64,953 |
|---|
Dots: score · lines: published uncertainty