ChessBench GitHub - Elo: leaderboard
Metric: Benchmark Elo. Source: chessbench-ai.github.io. 34 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 (Max) | 2640 |
| 2 | GPT-5.5 Pro | 2630 |
| 3 | Kimi K3 (Max) | 2590 |
| 4 | Gemini 3.1 Pro (Preview) (High) | 2570 |
| 5 | Qwen 3.8 Max | 2500 |
| 6 | Grok 4.6 (xHigh) | 2480 |
| 7 | GLM-5.3 Flash | 2450 |
| 8 | Claude Opus 5 (High) | 2400 |
| 9 | Gemma 4 | 2380 |
| 10 | Claude Fable 5.1 (Max) | 2370 |
| 11 | Gemini 3.6 Flash | 2300 |
| 12 | Gemini 3.8 Flash (Medium) | 2300 |
| 13 | GPT-4.5 | 2280 |
| 14 | Gemini 3.1 Pro (Preview) | 2240 |
| 15 | MiniMax-M3 | 2235 |
Interactive version: theaggregate.ai/benchmark?slug=chessbench-github-elo · How It Works · Data refreshed daily, snapshot 2026-09-19.