Arena AI Code: leaderboard
Crowdsourced Arena AI pairwise human-preference leaderboard for code generation and coding-assistant models.
Metric: Arena ELO (imported). Source: arena.ai. Source: benchmarklist.com. 120 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | claude-fable-5.1-max | 1765.37 | |
| 2 | qwen3.8-max-0902 | 1687.74 | |
| 3 | claude-opus-5-max | 1687.23 | |
| 4 | kimi-k3-max | 1673.82 | |
| 5 | qwen3.8-max | 1669.22 | |
| 6 | claude-opus-5-high | 1663 | |
| 7 | claude-opus-5-high | 1661.23 | |
| 8 | grok-4.6-high | 1629.06 | |
| 9 | claude-fable-5 | 1628.42 | |
| 10 | hy4-preview | 1625.78 |
Interactive version: theaggregate.ai/benchmark?slug=arena-ai-code · How It Works · Data refreshed daily, snapshot 2026-10-20.