SnakeBench: leaderboard
LLM Battle Snake arena where AI models play the classic Snake game competitively, testing strategic planning and decision-making.
Metric: TrueSkill Rating. Source: snakebench.com. Saturation forecast: Around October 2027. 322 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.1 Codex Max | 38.2 | #100 |
| 2 | Gemini 3 Pro (Preview) | 38 | #66 |
| 3 | GPT-6 Astra | 37.9 | #6 |
| 4 | GPT-5.5 | 37.4 | #31 |
| 5 | Gemini 3.1 Pro (Preview) | 37 | #58 |
| 6 | GPT-5.1 Codex Mini | 36.5 | #286 |
| 7 | O3 | 35.9 | #124 |
| 8 | Claude Opus 4.8 | 35.4 | #37 |
| 9 | GPT-5 | 35 | #98 |
| 10 | O4 Mini | 34.5 | #180 |
| 11 | Qwen 3.7 Max | 34.1 | #75 |
| 12 | Gemini 3.5 Flash | 33.8 | #60 |
| 13 | O4 Mini (High) | 33.7 | #180 (O4 Mini) |
| 14 | GPT-5.3 Codex | 33.4 | #69 |
| 15 | GPT-5 Codex | 33.3 | #127 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=snakebench · How It Works · Data refreshed daily, snapshot 2026-10-20.