A-Fantasia Chess
Games · 2025-04-26
A-Fantasia Chess asks for a Standard Algebraic Notation move after a randomly generated legal move sequence. A response is correct when it names any legal move in the resulting position; scores are error rates over 100 trials, so lower is better.
Top models (lower is better)
| Model | Score |
|---|---|
| GPT-4.5 | 3.0 |
| GPT-5 Chat | 11.0 |
| Gemini 2.0 Flash | 12.0 |
| GPT-4o (source unspecified) | 13.0 |
| GPT-4.1 | 13.0 |
| GPT-5.5 | 14.0 |
| Qwen3.7-Max (source unspecified) | 16.0 |
| GPT-5.1 | 16.0 |
| Opus 4.1 | 19.0 |
| Opus 4.6 | 20.0 |
| Opus 4.5 | 22.0 |
| GPT-5.6 Sol | 22.0 |
| GPT-5.6 Terra | 22.0 |
| Gemini 2.0 Flash-Lite | 22.0 |
| Sonnet 4.5 | 25.0 |
| GPT-5.6 Luna | 25.0 |
| Sonnet 4.6 | 26.0 |
| Gemini 3.1 Flash-Lite | 26.0 |
| Gemini 2.5 Flash | 27.0 |
| Grok 3 (API) | 28.0 |
| Opus 4 | 29.0 |
| Sonnet 4 | 29.0 |
| Kimi K2.6 | 29.0 |
| Gemini 3 Flash Preview | 30.0 |
| Opus 4.8 | 31.0 |
| GPT-5.4 | 33.0 |
| Claude 3.7 Sonnet | 33.0 |
| Qwen3.6 Plus (2026-04-02) | 33.0 |
| Llama 3.3 70B Instruct | 34.0 |
| Claude 3.5 Sonnet (Oct 2024) | 35.0 |
| Gemini 1.5 Pro (source unspecified) | 35.0 |
| Llama 3.1 405B Instruct | 38.0 |
| Opus 3 | 42.0 |
| GPT-5.2 | 42.0 |
| Qwen3 Max (rolling alias) | 43.0 |
| DeepSeek-V4-Pro | 44.0 |
| Qwen3.7-Plus (source unspecified) | 45.0 |
| DeepSeek-V3-0324 | 46.0 |
| Kimi K2 Instruct | 51.0 |
| GLM-5 | 52.0 |