← The Token Games

All Duels

90 duels between 10 frontier models — click any match to explore its puzzles.

Player AScorePlayer B
claude-opus-4-5-20251101 0 vs -2 claude-sonnet-4-5-20250929 view
claude-opus-4-5-20251101 -1 vs -1 deepseek-reasoner view
claude-opus-4-5-20251101 0 vs -1 gemini-2.5-flash view
claude-opus-4-5-20251101 1 vs -4 gemini-2.5-pro view
claude-opus-4-5-20251101 0 vs -1 gemini-3-pro-preview view
claude-opus-4-5-20251101 2 vs -2 gpt-5-mini-2025-08-07 view
claude-opus-4-5-20251101 1 vs -5 gpt-5.2-2025-12-11 view
claude-opus-4-5-20251101 -3 vs 1 gpt-5.2-pro-2025-12-11 view
claude-opus-4-5-20251101 0 vs -4 grok-4 view
claude-sonnet-4-5-20250929 0 vs 1 claude-opus-4-5-20251101 view
claude-sonnet-4-5-20250929 0 vs 1 deepseek-reasoner view
claude-sonnet-4-5-20250929 0 vs -1 gemini-2.5-flash view
claude-sonnet-4-5-20250929 0 vs -5 gemini-2.5-pro view
claude-sonnet-4-5-20250929 0 vs 3 gemini-3-pro-preview view
claude-sonnet-4-5-20250929 0 vs 0 gpt-5-mini-2025-08-07 view
claude-sonnet-4-5-20250929 1 vs -5 gpt-5.2-2025-12-11 view
claude-sonnet-4-5-20250929 -2 vs 5 gpt-5.2-pro-2025-12-11 view
claude-sonnet-4-5-20250929 0 vs 0 grok-4 view
deepseek-reasoner 3 vs 2 claude-opus-4-5-20251101 view
deepseek-reasoner 1 vs -1 claude-sonnet-4-5-20250929 view
deepseek-reasoner 0 vs -5 gemini-2.5-flash view
deepseek-reasoner -1 vs -5 gemini-2.5-pro view
deepseek-reasoner -1 vs -2 gemini-3-pro-preview view
deepseek-reasoner -1 vs -2 gpt-5-mini-2025-08-07 view
deepseek-reasoner 2 vs -5 gpt-5.2-2025-12-11 view
deepseek-reasoner -1 vs 0 gpt-5.2-pro-2025-12-11 view
deepseek-reasoner 0 vs -1 grok-4 view
gemini-2.5-flash -1 vs 1 claude-opus-4-5-20251101 view
gemini-2.5-flash -4 vs 0 claude-sonnet-4-5-20250929 view
gemini-2.5-flash 0 vs -1 deepseek-reasoner view
gemini-2.5-flash -1 vs -3 gemini-2.5-pro view
gemini-2.5-flash -2 vs 2 gemini-3-pro-preview view
gemini-2.5-flash -3 vs 0 gpt-5-mini-2025-08-07 view
gemini-2.5-flash -4 vs -5 gpt-5.2-2025-12-11 view
gemini-2.5-flash -1 vs 4 gpt-5.2-pro-2025-12-11 view
gemini-2.5-flash -1 vs 0 grok-4 view
gemini-2.5-pro -5 vs 0 claude-opus-4-5-20251101 view
gemini-2.5-pro -5 vs -2 claude-sonnet-4-5-20250929 view
gemini-2.5-pro -4 vs -3 deepseek-reasoner view
gemini-2.5-pro -3 vs -5 gemini-2.5-flash view
gemini-2.5-pro -5 vs 1 gemini-3-pro-preview view
gemini-2.5-pro -4 vs 0 gpt-5-mini-2025-08-07 view
gemini-2.5-pro -5 vs -5 gpt-5.2-2025-12-11 view
gemini-2.5-pro -5 vs 3 gpt-5.2-pro-2025-12-11 view
gemini-2.5-pro -4 vs 1 grok-4 view
gemini-3-pro-preview 0 vs 0 claude-opus-4-5-20251101 view
gemini-3-pro-preview 3 vs -2 claude-sonnet-4-5-20250929 view
gemini-3-pro-preview 2 vs -1 deepseek-reasoner view
gemini-3-pro-preview -2 vs -2 gemini-2.5-flash view
gemini-3-pro-preview -1 vs -5 gemini-2.5-pro view
gemini-3-pro-preview -3 vs 0 gpt-5-mini-2025-08-07 view
gemini-3-pro-preview 0 vs -5 gpt-5.2-2025-12-11 view
gemini-3-pro-preview -3 vs 0 gpt-5.2-pro-2025-12-11 view
gemini-3-pro-preview -1 vs -1 grok-4 view
gpt-5-mini-2025-08-07 -1 vs 0 claude-opus-4-5-20251101 view
gpt-5-mini-2025-08-07 -3 vs -1 claude-sonnet-4-5-20250929 view
gpt-5-mini-2025-08-07 -2 vs -2 deepseek-reasoner view
gpt-5-mini-2025-08-07 2 vs -4 gemini-2.5-flash view
gpt-5-mini-2025-08-07 -2 vs -5 gemini-2.5-pro view
gpt-5-mini-2025-08-07 -1 vs -4 gemini-3-pro-preview view
gpt-5-mini-2025-08-07 4 vs -5 gpt-5.2-2025-12-11 view
gpt-5-mini-2025-08-07 -1 vs 2 gpt-5.2-pro-2025-12-11 view
gpt-5-mini-2025-08-07 0 vs -3 grok-4 view
gpt-5.2-2025-12-11 -5 vs 1 claude-opus-4-5-20251101 view
gpt-5.2-2025-12-11 -5 vs 0 claude-sonnet-4-5-20250929 view
gpt-5.2-2025-12-11 -5 vs 2 deepseek-reasoner view
gpt-5.2-2025-12-11 -5 vs -5 gemini-2.5-flash view
gpt-5.2-2025-12-11 -5 vs -5 gemini-2.5-pro view
gpt-5.2-2025-12-11 -3 vs -1 gemini-3-pro-preview view
gpt-5.2-2025-12-11 -5 vs 2 gpt-5-mini-2025-08-07 view
gpt-5.2-2025-12-11 -5 vs 4 gpt-5.2-pro-2025-12-11 view
gpt-5.2-2025-12-11 -5 vs 0 grok-4 view
gpt-5.2-pro-2025-12-11 0 vs -4 claude-opus-4-5-20251101 view
gpt-5.2-pro-2025-12-11 3 vs -4 claude-sonnet-4-5-20250929 view
gpt-5.2-pro-2025-12-11 4 vs -5 deepseek-reasoner view
gpt-5.2-pro-2025-12-11 1 vs -5 gemini-2.5-flash view
gpt-5.2-pro-2025-12-11 -1 vs -5 gemini-2.5-pro view
gpt-5.2-pro-2025-12-11 1 vs -3 gemini-3-pro-preview view
gpt-5.2-pro-2025-12-11 -1 vs -4 gpt-5-mini-2025-08-07 view
gpt-5.2-pro-2025-12-11 -1 vs -5 gpt-5.2-2025-12-11 view
gpt-5.2-pro-2025-12-11 0 vs -3 grok-4 view
grok-4 -1 vs 0 claude-opus-4-5-20251101 view
grok-4 0 vs -1 claude-sonnet-4-5-20250929 view
grok-4 0 vs -1 deepseek-reasoner view
grok-4 1 vs -2 gemini-2.5-flash view
grok-4 1 vs -3 gemini-2.5-pro view
grok-4 0 vs 1 gemini-3-pro-preview view
grok-4 -2 vs -2 gpt-5-mini-2025-08-07 view
grok-4 -2 vs -5 gpt-5.2-2025-12-11 view
grok-4 -1 vs 1 gpt-5.2-pro-2025-12-11 view

The Token Games · Henniger & Poesia · Harvard University · 2026