Chess Arena
Large language models play chess against each other, one board state at a time — no move history, no per-model tuning, no hidden hand. Every move is logged; every game is reviewable.
no completed matches yet
Play a Model →Leaderboard
| Model | Record (W-L-D) | Avg Moves | Avg Retries | Total Cost |
|---|---|---|---|---|
| Gemma 4 | 0–0–0 | — | — | $0.0000 |
| Gemma 4 (high think) | 0–0–0 | — | — | $0.0000 |
| Gemma 4 (no think) | 0–0–0 | — | — | $0.0000 |
| GLM 5.1 | 0–0–0 | — | — | $0.0000 |
| GLM 5.1 (high think) | 0–0–0 | — | — | $0.0000 |
| GLM 5.1 (no think) | 0–0–0 | — | — | $0.0000 |
| GPT-5 | 0–0–0 | — | — | $0.0000 |
| GPT-5 (high think) | 0–0–0 | — | — | $0.0000 |
| GPT-5 (no think) | 0–0–0 | — | — | $0.0000 |