Match statistics
Human vs Model
11
Total
7
Human wins
1
Model wins
0
Draws
3
Aborted*
0
In progress
| Model | W | L | D | A |
|---|---|---|---|---|
| OpenAI · gpt-6-astra | 1 | 0 | 0 | 0 |
| DeepSeek · deepseek-v4-flash | 0 | 2 | 0 | 3 |
| OpenAI · gpt-5.6-luna | 0 | 5 | 0 | 0 |
| Game | Human wins | Model wins | Draws | Aborted* |
|---|---|---|---|---|
| Chess | 1 | 0 | 0 | 1 |
| Go | 1 | 0 | 0 | 0 |
| Gomoku | 3 | 1 | 0 | 2 |
| Othello | 1 | 0 | 0 | 0 |
| Xiangqi | 1 | 0 | 0 | 0 |
Model vs Model
23
Total
21
Decided
1
Draws
1
Aborted*
0
In progress
Per-model record
| Model | W | L | D | A |
|---|---|---|---|---|
| OpenAI · gpt-6-astra | 9 | 1 | 0 | 1 |
| Anthropic · claude-fable-5-1 | 4 | 5 | 0 | 0 |
| DeepSeek · deepseek-flash | 2 | 0 | 0 | 0 |
| Moonshot · kimi-k3 | 2 | 4 | 0 | 0 |
| Google · gemini-3.8-flash | 1 | 1 | 0 | 0 |
| Meta · muse-spark-1.2 | 1 | 1 | 0 | 0 |
| OpenAI · gpt-5.6-luna | 1 | 0 | 1 | 0 |
| OpenAI · gpt-5.6-sol | 1 | 0 | 0 | 0 |
| Anthropic · claude-sonnet-5 | 0 | 1 | 1 | 0 |
| OpenAI · gpt-5.6-terra | 0 | 1 | 0 | 0 |
| Zhipu · glm-5.3 | 0 | 4 | 0 | 0 |
| xAI · grok-4.6 | 0 | 2 | 0 | 0 |
| xAI · grok-4.7 | 0 | 1 | 0 | 1 |
By game
Chess: 16Gomoku: 2Othello: 1Xiangqi: 4
Model output failures
Persisted failed attempts from move history. Illegal columns are counted strikes, split by reasoning level. Soft failures (refusals, safety / content filters, etc.) do not count as strikes.
| Model | Total | Illegal (off) | Illegal (low) | Illegal (medium) | Illegal (high) | Soft |
|---|---|---|---|---|---|---|
| Zhipu · glm-5.3 | 6 | — | 6 | — | — | — |
| DeepSeek · deepseek-v4-flash | 6 | 3 | 3 | — | — | — |
| Anthropic · claude-fable-5-1 | 3 | — | 3 | — | — | — |
| Anthropic · claude-sonnet-5 | 1 | — | 1 | — | — | — |
Model think time
Average time from starting an LLM request to accepting a legal move, split by reasoning level.
| Model | Moves (off) | Avg (off) | Moves (low) | Avg (low) | Moves (medium) | Avg (medium) | Moves (high) | Avg (high) |
|---|---|---|---|---|---|---|---|---|
| Anthropic · claude-sonnet-5 | — | — | 59 | 1.7 s | — | — | — | — |
| Google · gemini-3.8-flash | — | — | 62 | 2.3 s | — | — | — | — |
| OpenAI · gpt-5.6-luna | — | — | 58 | 6.6 s | — | — | — | — |
| OpenAI · gpt-6-astra | — | — | 321 | 11.0 s | — | — | — | — |
| Anthropic · claude-fable-5-1 | — | — | 265 | 11.3 s | — | — | — | — |
| OpenAI · gpt-5.6-sol | — | — | 39 | 11.7 s | — | — | — | — |
| OpenAI · gpt-5.6-terra | — | — | 13 | 14.4 s | — | — | — | — |
| Zhipu · glm-5.3 | — | — | 81 | 14.5 s | — | — | — | — |
| xAI · grok-4.6 | — | — | 63 | 15.2 s | — | — | — | — |
| Meta · muse-spark-1.2 | — | — | 35 | 21.9 s | — | — | — | — |
| Moonshot · kimi-k3 | — | — | 177 | 26.1 s | — | — | — | — |
| xAI · grok-4.7 | — | — | 35 | 38.8 s | — | — | — | — |
| DeepSeek · deepseek-v4-flash | 3 | 1.0 s | 1 | 11.5 s | 2 | 9.6 s | 2 | 55.4 s |
| DeepSeek · deepseek-flash | — | — | 38 | 1 m 4 s | — | — | — | — |
Per-player record (Human vs Model)
By the name you entered. Players without a name are counted in one Human row.
| Player | Games | W | L | D | A |
|---|---|---|---|---|---|
| Human | 11 | 7 | 1 | 0 | 3 |