aiaimodel.cards
← All benchmarks

GPQA Diamond (no tools)

Graduate-level science multiple-choice questions written to be hard to look up.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Gemini 3.1 Flash-Lite · Googlebest value
    86.9%
  2. 2gpt-oss-120b · OpenAIbest value
    80.1%
  3. 3gpt-oss-20b · OpenAIbest value
    71.5%

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Gemini 3.1 Flash-Lite p.4 (table also on the model card web page)

    Results as of March 2026, as published in Google's model card. Settings as labelled by Google: 3.1 Flash-Lite High, 2.5 Flash and 2.5 Flash-Lite Dynamic, GPT-5 mini High, Claude 4.5 Haiku Extended Thinking, Grok 4.1 Fast Reasoning. GPT-5 mini CharXiv value is with python.

  • gpt-oss-120b Figures 1, 2, 4; Table 2

    OpenAI's own pass@1 results at reasoning level high. AIME and HLE with tools; Codeforces with a terminal tool. o4-mini and o3 values read from OpenAI's charts.

  • gpt-oss-20b Figures 1, 2, 4; Table 2

    OpenAI's own pass@1 results at reasoning level high. AIME and HLE with tools; Codeforces with a terminal tool. o3-mini was run without tools on AIME, HLE and Codeforces, so those cells are left empty. Baseline values read from OpenAI's charts.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.