Science
GPQA Diamond
Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better.
Rankings
Higher is better194.4%294.3%394.2%493.6%592.8%692.4%791.9%891.3%990.5%1090.4%1190.1%1289.9%1389.5%1488.1%1488.1%1687.6%1787%1886.4%1983.4%2080.9%2179.6%2275.4%2375.1%2375.1%2573%2668%2759.4%2850.4%2941.6%3040.4%3133.3%
GPT-5.4-Pro
OpenAI · Mar 5 2026
Gemini 3.1 Pro
Google · Feb 19 2026
Claude Opus 4.7
Anthropic · Apr 16 2026
GPT-5.5
OpenAI · Apr 23 2026
GPT-5.4
OpenAI · Mar 5 2026
GPT-5.2
OpenAI · Dec 11 2025
Gemini 3.0 Pro
Google · Nov 18 2025
Claude Opus 4.6
Anthropic · Feb 5 2026
Kimi K2.6
Moonshot AI · Apr 21 2026
Gemini 3.0 Flash
Google · Dec 17 2025
DeepSeek-V4-Pro
DeepSeek · Apr 24 2026
Claude Sonnet 4.6
Anthropic · Feb 17 2026
Muse Spark
Meta · Apr 8 2026
GPT-5.1
OpenAI · Nov 12 2025
DeepSeek-V4-Flash
DeepSeek · Apr 24 2026
Kimi K2.5
Moonshot AI · Jan 27 2026
Claude Opus 4.5
Anthropic · Nov 24 2025
Gemini 2.5 Pro
Google · Mar 25 2025
Claude Sonnet 4.5
Anthropic · Sep 29 2025
Claude Opus 4.1
Anthropic · Aug 5 2025
Claude Opus 4
Anthropic · May 22 2025
Claude Sonnet 4
Anthropic · May 22 2025
Kimi K2
Moonshot AI · Jul 11 2025
Kimi K2 (0905)
Moonshot AI · Sep 5 2025
Claude Haiku 4.5
Anthropic · Oct 15 2025
Claude 3.7 Sonnet
Anthropic · Feb 24 2025
Claude 3.5 Sonnet
Anthropic · Jun 20 2024
Claude 3 Opus
Anthropic · Mar 4 2024
Claude 3.5 Haiku
Anthropic · Oct 22 2024
Claude 3 Sonnet
Anthropic · Mar 4 2024
Claude 3 Haiku
Anthropic · Mar 4 2024