Vibeleaderboard
← Benchmark guide

Model results

GPQA Diamond

Choose a shared evaluation protocol to compare model scores. Versions, sources and configurations stay separate.

Atlas snapshot captured 10/19/2026, 10:20:28 AM UTC. This is an import timestamp, not the evaluation date. Coverage may be incomplete.

Reviewed recommendation and limits

Use as a compact reference for expert-written science questions, alongside evidence that matches the scientific work you intend to do.

Original GPQA Diamond subset. Preserve prompt, answer extraction, tools, sampling and aggregation. Keep main, extended and third-party corrected subsets separate.. Imported groups below may use different protocols; this review does not certify each run. The static 198-question subset has limited statistical power and exposure risk. Its expert filtering does not make it representative of scientific practice. A Google-proof collection process does not prove present-day contamination resistance.

Publisher results GPQA paper
← All evaluation groups

GPQA

Source
Artificial Analysis
Version
Not reported
Measurement date
Not reported
Metric
% · Higher is better
Evaluation scope
Graduate-level science questions; observed 2026-10-19; result date not supplied

Exact score ties share a position. Positions do not establish statistical significance. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

  1. #1GPT-6 Astra

    openai/gpt-6-astra-20260903

    96.1%

    View modelEvidence independent

  2. #2Gemini 3.8 Flash

    google/gemini-3.8-flash-20260902

    95.3%

    View modelEvidence independent

  3. #3Grok 4.6

    x-ai/grok-4.6-20260810

    94.9%

    View modelEvidence independent

  4. #4Gemini 3.7 Flash

    google/gemini-3.7-flash-20260813

    94.5%

    View modelEvidence independent

  5. #5GPT-5.6 Sol

    openai/gpt-5.6-sol-20260709

    94.1%

    View modelEvidence independent

  6. #6Claude Fable 5.1

    anthropic/claude-fable-5-1

    93.7%

    View modelEvidence independent

  7. #7Kimi K3

    moonshotai/kimi-k3-20260715

    93.5%

    View modelEvidence independent

  8. #7Muse Spark 1.3

    meta/muse-spark-1.3-20260902

    93.5%

    View modelEvidence independent

  9. #9Claude Opus 5

    anthropic/claude-opus-5

    93.2%

    View modelEvidence independent

  10. #10MiniMax M3

    minimax/minimax-m3-20260531

    92.9%

    View modelEvidence independent

  11. #11DeepSeek V4 Pro 0813

    deepseek/deepseek-v4-pro-20260813

    92.8%

    View modelEvidence independent

  12. #11Qwen3.8 Max (0902)

    qwen/qwen3.8-max-20260902

    92.8%

    View modelEvidence independent

  13. #13Claude Fable 5

    anthropic/claude-fable-5

    92.6%

    View modelEvidence independent

  14. #14GPT-5.6 Terra

    openai/gpt-5.6-terra-20260709

    92.5%

    View modelEvidence independent

  15. #15GLM 5.3

    z-ai/glm-5.3-20260816

    91.7%

    View modelEvidence independent

  16. #16GLM 5.3 Flash

    z-ai/glm-5.3-flash-20260826

    91.2%

    View modelEvidence independent

  17. #17GPT-5.6 Luna

    openai/gpt-5.6-luna-20260709

    91.1%

    View modelEvidence independent

  18. #18Qwen3.8 27B

    qwen/qwen3.8-27b-20260814

    90.5%

    View modelEvidence independent

  19. #19Muse Spark 1.2

    meta/muse-spark-1.2-20260805

    90.4%

    View modelEvidence independent

  20. #20Hy3

    tencent/hy3-20260706

    89.7%

    View modelEvidence independent

  21. #20Hy3 preview

    tencent/hy3-preview-20260421

    89.7%

    View modelEvidence independent

  22. #22Muse Spark Open

    meta/muse-spark-open-20261015

    89%

    View modelEvidence independent

  23. #23Inkling

    thinkingmachines/inkling-20260715

    87.2%

    View modelEvidence independent

  24. #24Nemotron 3 Ultra

    nvidia/nemotron-3-ultra-550b-a55b-20260604

    86.7%

    View modelEvidence independent

  25. #25MiMo-V2.5-Pro

    xiaomi/mimo-v2.5-pro-20260422

    86.6%

    View modelEvidence independent

  26. #26MiMo-V2.5

    xiaomi/mimo-v2.5-20260422

    84.9%

    View modelEvidence independent

  27. #27Gemini 3.5 Flash Lite

    google/gemini-3.5-flash-lite-20260721

    83.8%

    View modelEvidence independent

  28. #28gpt-oss-120b

    openai/gpt-oss-120b

    78.2%

    View modelEvidence independent

  29. #29Command A

    cohere/command-a-03-2025

    76.1%

    View modelEvidence independent

  30. #29Command A+

    cohere/command-a-plus-05-2026

    76.1%

    View modelEvidence independent

  31. #31Mistral Medium 3

    mistralai/mistral-medium-3

    74.8%

    View modelEvidence independent

  32. #31Mistral Medium 3.1

    mistralai/mistral-medium-3.1

    74.8%

    View modelEvidence independent

  33. #31Mistral Medium 3.5

    mistralai/mistral-medium-3.5-20260430

    74.8%

    View modelEvidence independent

  34. #34Nemotron 3.5 Lightning

    nvidia/nemotron-3.5-lightning-20260807

    74.3%

    View modelEvidence independent

  35. #35MiMo-V2-Flash

    xiaomi/mimo-v2-flash

    65.6%

    View modelEvidence independent

  36. #36Phi 4

    microsoft/phi-4

    57.5%

    View modelEvidence independent
GPQA Diamond model results | VibeLeaderboard