90 duels between 10 frontier models — click any match to explore its puzzles.
| Player A | Score | Player B | |
|---|---|---|---|
| claude-haiku-4-5 | 0 vs 0 | claude-opus-4-7 | view |
| claude-haiku-4-5 | 0 vs 0 | claude-sonnet-4-6 | view |
| claude-haiku-4-5 | -1 vs -1 | deepseek-v3.2-thinking | view |
| claude-haiku-4-5 | 0 vs 0 | gemini-3-flash-preview | view |
| claude-haiku-4-5 | 0 vs 0 | gemini-3.1-pro-preview | view |
| claude-haiku-4-5 | 0 vs -1 | gpt-5.4-mini | view |
| claude-haiku-4-5 | 0 vs 5 | gpt-5.5 | view |
| claude-haiku-4-5 | 0 vs 0 | grok-4-fast-reasoning | view |
| claude-haiku-4-5 | 0 vs -1 | grok-4.20-0309-reasoning | view |
| claude-opus-4-7 | 1 vs 0 | claude-haiku-4-5 | view |
| claude-opus-4-7 | 0 vs -1 | claude-sonnet-4-6 | view |
| claude-opus-4-7 | 0 vs -2 | deepseek-v3.2-thinking | view |
| claude-opus-4-7 | 1 vs -4 | gemini-3-flash-preview | view |
| claude-opus-4-7 | 1 vs -1 | gemini-3.1-pro-preview | view |
| claude-opus-4-7 | 1 vs -4 | gpt-5.4-mini | view |
| claude-opus-4-7 | 0 vs 0 | gpt-5.5 | view |
| claude-opus-4-7 | 0 vs 0 | grok-4-fast-reasoning | view |
| claude-opus-4-7 | 0 vs 0 | grok-4.20-0309-reasoning | view |
| claude-sonnet-4-6 | 0 vs 0 | claude-haiku-4-5 | view |
| claude-sonnet-4-6 | 0 vs 0 | claude-opus-4-7 | view |
| claude-sonnet-4-6 | 1 vs 0 | deepseek-v3.2-thinking | view |
| claude-sonnet-4-6 | 0 vs 4 | gemini-3-flash-preview | view |
| claude-sonnet-4-6 | 0 vs 1 | gemini-3.1-pro-preview | view |
| claude-sonnet-4-6 | 0 vs -1 | gpt-5.4-mini | view |
| claude-sonnet-4-6 | 0 vs 1 | gpt-5.5 | view |
| claude-sonnet-4-6 | 0 vs -1 | grok-4-fast-reasoning | view |
| claude-sonnet-4-6 | 1 vs 0 | grok-4.20-0309-reasoning | view |
| deepseek-v3.2-thinking | -2 vs -1 | claude-haiku-4-5 | view |
| deepseek-v3.2-thinking | 1 vs 1 | claude-opus-4-7 | view |
| deepseek-v3.2-thinking | -1 vs 0 | claude-sonnet-4-6 | view |
| deepseek-v3.2-thinking | -1 vs 0 | gemini-3-flash-preview | view |
| deepseek-v3.2-thinking | -1 vs 4 | gemini-3.1-pro-preview | view |
| deepseek-v3.2-thinking | -1 vs -1 | gpt-5.4-mini | view |
| deepseek-v3.2-thinking | -2 vs 3 | gpt-5.5 | view |
| deepseek-v3.2-thinking | 0 vs -1 | grok-4-fast-reasoning | view |
| deepseek-v3.2-thinking | 4 vs 5 | grok-4.20-0309-reasoning | view |
| gemini-3-flash-preview | -4 vs 0 | claude-haiku-4-5 | view |
| gemini-3-flash-preview | -2 vs 4 | claude-opus-4-7 | view |
| gemini-3-flash-preview | -5 vs 2 | claude-sonnet-4-6 | view |
| gemini-3-flash-preview | -3 vs -4 | deepseek-v3.2-thinking | view |
| gemini-3-flash-preview | -2 vs 2 | gemini-3.1-pro-preview | view |
| gemini-3-flash-preview | 0 vs -1 | gpt-5.4-mini | view |
| gemini-3-flash-preview | -4 vs 2 | gpt-5.5 | view |
| gemini-3-flash-preview | 0 vs -3 | grok-4-fast-reasoning | view |
| gemini-3-flash-preview | -5 vs 2 | grok-4.20-0309-reasoning | view |
| gemini-3.1-pro-preview | 1 vs 0 | claude-haiku-4-5 | view |
| gemini-3.1-pro-preview | -1 vs 0 | claude-opus-4-7 | view |
| gemini-3.1-pro-preview | -1 vs 0 | claude-sonnet-4-6 | view |
| gemini-3.1-pro-preview | 5 vs -4 | deepseek-v3.2-thinking | view |
| gemini-3.1-pro-preview | 0 vs 1 | gemini-3-flash-preview | view |
| gemini-3.1-pro-preview | 0 vs -1 | gpt-5.4-mini | view |
| gemini-3.1-pro-preview | 1 vs 1 | gpt-5.5 | view |
| gemini-3.1-pro-preview | 2 vs 0 | grok-4-fast-reasoning | view |
| gemini-3.1-pro-preview | 0 vs -3 | grok-4.20-0309-reasoning | view |
| gpt-5.4-mini | 0 vs -2 | claude-haiku-4-5 | view |
| gpt-5.4-mini | -1 vs 1 | claude-opus-4-7 | view |
| gpt-5.4-mini | 2 vs 0 | claude-sonnet-4-6 | view |
| gpt-5.4-mini | -1 vs -3 | deepseek-v3.2-thinking | view |
| gpt-5.4-mini | 1 vs -3 | gemini-3-flash-preview | view |
| gpt-5.4-mini | -1 vs 0 | gemini-3.1-pro-preview | view |
| gpt-5.4-mini | -1 vs 1 | gpt-5.5 | view |
| gpt-5.4-mini | 1 vs 1 | grok-4-fast-reasoning | view |
| gpt-5.4-mini | 1 vs 0 | grok-4.20-0309-reasoning | view |
| gpt-5.5 | 2 vs 0 | claude-haiku-4-5 | view |
| gpt-5.5 | 1 vs 1 | claude-opus-4-7 | view |
| gpt-5.5 | 1 vs 0 | claude-sonnet-4-6 | view |
| gpt-5.5 | -1 vs -1 | deepseek-v3.2-thinking | view |
| gpt-5.5 | 2 vs -4 | gemini-3-flash-preview | view |
| gpt-5.5 | 0 vs 0 | gemini-3.1-pro-preview | view |
| gpt-5.5 | 0 vs -1 | gpt-5.4-mini | view |
| gpt-5.5 | 4 vs -3 | grok-4-fast-reasoning | view |
| gpt-5.5 | 0 vs -1 | grok-4.20-0309-reasoning | view |
| grok-4-fast-reasoning | 0 vs 0 | claude-haiku-4-5 | view |
| grok-4-fast-reasoning | -3 vs -1 | claude-opus-4-7 | view |
| grok-4-fast-reasoning | 0 vs 1 | claude-sonnet-4-6 | view |
| grok-4-fast-reasoning | -1 vs -2 | deepseek-v3.2-thinking | view |
| grok-4-fast-reasoning | 1 vs 3 | gemini-3-flash-preview | view |
| grok-4-fast-reasoning | -1 vs 4 | gemini-3.1-pro-preview | view |
| grok-4-fast-reasoning | -3 vs 2 | gpt-5.4-mini | view |
| grok-4-fast-reasoning | -2 vs 4 | gpt-5.5 | view |
| grok-4-fast-reasoning | 0 vs 0 | grok-4.20-0309-reasoning | view |
| grok-4.20-0309-reasoning | -1 vs 0 | claude-haiku-4-5 | view |
| grok-4.20-0309-reasoning | -3 vs 1 | claude-opus-4-7 | view |
| grok-4.20-0309-reasoning | 2 vs 4 | claude-sonnet-4-6 | view |
| grok-4.20-0309-reasoning | 0 vs 0 | deepseek-v3.2-thinking | view |
| grok-4.20-0309-reasoning | 3 vs -2 | gemini-3-flash-preview | view |
| grok-4.20-0309-reasoning | 2 vs 1 | gemini-3.1-pro-preview | view |
| grok-4.20-0309-reasoning | 1 vs -2 | gpt-5.4-mini | view |
| grok-4.20-0309-reasoning | 0 vs 1 | gpt-5.5 | view |
| grok-4.20-0309-reasoning | 0 vs -3 | grok-4-fast-reasoning | view |
The Token Games · Henniger & Poesia · Harvard University · 2026