| # | Model | Arena score | Samples |
|---|---|---|---|
| 21 | qwen3.8-maxAlibaba | 1,481±6 | 16,263 |
| 22 |  gemini-3.6-flash-highGoogle | 1,479±5 | 25,953 |
| 23 | gemini-3.5-flash-highGoogle | 1,478±4 | 37,808 |
| 24 |  gpt-6-astra-maxOpenAI | 1,478±13 | 2,059 |
| 25 | gpt-5.4-highOpenAI | 1,476±4 | 60,538 |
| 26 | 1,476±4 | 34,173 |
| 27 | gpt-5.5OpenAI | 1,476±4 | 66,349 |
|---|
| 28 |  glm-5.3-flashZ.ai | 1,475±7 | 9,479 |
|---|
| 29 | grok-4.20-beta1SpaceXAI | 1,475±5 | 26,595 |
|---|
| 30 | gemini-3.5-flash-mediumGoogle | 1,474±5 | 36,151 |
|---|
| 31 | gpt-5.5-instantOpenAI | 1,474±5 | 25,827 |
|---|
| 32 | gemini-3-flashGoogle | 1,474±4 | 30,239 |
|---|
| 33 | qwen3.7-max-previewAlibaba | 1,473±10 | 3,705 |
|---|
| 34 |  claude-opus-4-5-20251101-high-32kAnthropic | 1,473±4 | 36,240 |
|---|
| 35 | claude-opus-4-8Anthropic | 1,473±4 | 52,970 |
|---|
| 36 | claude-sonnet-4-6Anthropic | 1,473±4 | 66,208 |
|---|
| 37 | glm-5.2-maxZ.ai | 1,472±5 | 36,471 |
|---|
| 38 | grok-4.20-beta-0309-reasoningSpaceXAI | 1,471±4 | 62,168 |
|---|
| 39 | grok-4.20-multi-agent-beta-0309SpaceXAI | 1,470±4 | 60,775 |
|---|
| 40 | claude-opus-4-5-20251101Anthropic | 1,469±3 | 70,012 |
|---|
Dots: score · lines: published uncertainty