Benchmarks · reasoning
LLM Chess
LLMs play chess through a text tool interface against a random-move player and the Komodo Dragon engine at set skill levels; the value is the Elo estimated from those games.
102 models · 144 results · updated 2026-10-21 · maxim-saplin.github.io ↗
Leaderboard
GPT-6 Astra high
More from this benchmark
8 published measuresScore vs any published measure8 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | ||||
|---|---|---|---|---|
| GPT-6 Astra high | 109.7 | 67.0 | 20.0 | 47.0 |
| GPT-6.1 Sol xhigh | 161.9 | 30.0 | 10.0 | 20.0 |
| Gemini 4 Argon high | 130.6 | 40.0 | 10.0 | 30.0 |
| Opus 5.5 high, fallback | 112.7 | 64.0 | 10.0 | 54.0 |
| GPT-5.6 Sol xhigh | 156.4 | 30.0 | 10.0 | 20.0 |
| Gemini 3.8 Flash high | 170.5 | 32.0 | 12.0 | 20.0 |
| GPT-6 Sol xhigh | 175.7 | 21.0 | 0 | 21.0 |
| GPT-5.5 medium | 147.6 | 33.0 | 10.0 | 23.0 |
| Gemini 3.1 Pro high | 147.7 | 67.0 | 10.0 | 57.0 |
| GPT-5.4 high | 133.5 | 66.0 | 24.0 | 42.0 |
| Mercury 3 high | 126.3 | 62.0 | 20.0 | 42.0 |
| GPT-5.6 Terra xhigh | 152.4 | 32.0 | 11.0 | 21.0 |
| Opus 5 high | 127.4 | 53.0 | 10.0 | 43.0 |
| GPT-6 Luna max | 144.3 | 35.0 | 10.0 | 25.0 |
| Sonnet 5.5 high, fallback | 120.4 | 74.0 | 25.0 | 49.0 |
| GPT-5.6 Luna xhigh | 116.8 | 51.0 | 10.0 | 41.0 |
| Haiku 5.5 high | 107.7 | 60.0 | 10.0 | 50.0 |
| Gemini 3.7 Flash high | 146.5 | 32.0 | 9 | 23.0 |
| Gemini 3.6 Flash high | 109.0 | 59.0 | 10.0 | 49.0 |
| DeepSeek V4.1 Flash 2026 high | 181.0 | 25.0 | 10.0 | 15.0 |
| GPT-5 medium | 115.4 | 61.0 | 18.0 | 43.0 |
| GPT-5.2 medium | 118.6 | 57.0 | 15.0 | 42.0 |
| GPT-5.3 Codex medium | 147.5 | 53.0 | 15.0 | 38.0 |
| Gemini 3 Pro high | 143.9 | 53.0 | 27.0 | 26.0 |
| Gemini 3.5 Flash medium | 118.8 | 61.0 | 17.0 | 44.0 |
All results
144 rows| # | Model | Effort | LLM Chess |
|---|---|---|---|
| 1 | GPT-6 Astra | high | 1614 ±110 |
| 2 | GPT-6.1 Sol | xhigh | 1612 ±162 |
| 3 | Gemini 4 Argon | high | 1609 ±131 |
| 4 | GPT-6 Astra | xhigh | 1579 ±148 |
| 5 | Opus 5.5 | high | 1575 ±113 |
| 6 | GPT-5.6 Sol | xhigh | 1550 ±156 |
| 7 | GPT-5.6 Sol | high | 1548 ±137 |
| 8 | Gemini 3.8 Flash | high | 1546 ±170 |
| 9 | GPT-6 Sol | xhigh | 1545 ±176 |
| 10 | GPT-5.5 | medium | 1532 ±148 |
| 11 | GPT-5.5 | high | 1528 ±179 |
| 12 | Gemini 3.8 Flash | medium | 1527 ±135 |
| 13 | Gemini 3.1 Pro | high | 1511 ±148 |
| 14 | GPT-6.1 Sol | high | 1454 ±148 |
| 15 | GPT-5.4 | high | 1322 ±133 |
| 16 | GPT-6 Sol | high | 1310 ±122 |
| 17 | Mercury 3 | high | 1308 ±126 |
| 18 | GPT-5.6 Terra | xhigh | 1293 ±152 |
| 19 | Opus 5 | high | 1285 ±127 |
| 20 | GPT-5.4 | medium | 1253 ±112 |
| 21 | GPT-5.6 Sol | medium | 1240 ±165 |
| 22 | GPT-6 Luna | max | 1219 ±144 |
| 23 | GPT-5.6 Terra | high | 1211 ±131 |
| 24 | Sonnet 5.5 | high | 1195 ±120 |
| 25 | GPT-5.6 Luna | xhigh | 1185 ±117 |
| 26 | Haiku 5.5 | high | 1143 ±108 |
| 27 | Gemini 3.7 Flash | high | 1107 ±146 |
| 28 | Gemini 3.6 Flash | high | 1094 ±109 |
| 29 | DeepSeek V4.1 Flash 2026 | high | 1087 ±181 |
| 30 | GPT-5 | medium | 1086 ±115 |