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
Hugging Face OpenEvals
Version
Not reported
Measurement date
Not reported
Metric
% · Higher is better
Evaluation scope
Latest published official benchmark result aggregated by OpenEvals · result date not supplied

Positions are withheld because at least one result lacks explicit independent provenance or a documented evaluation scope. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

  1. Qwen3.5 397B A17B

    qwen/qwen3.5-397b-a17b-20260216

    88.4%

    View modelEvidence Evidence class unreported

  2. Kimi K2.5

    moonshotai/kimi-k2.5-0127

    87.6%

    View modelEvidence Evidence class unreported

  3. Qwen3.5-122B-A10B

    qwen/qwen3.5-122b-a10b-20260224

    86.6%

    View modelEvidence Evidence class unreported

  4. GLM 5

    z-ai/glm-5-20260211

    86%

    View modelEvidence Evidence class unreported

  5. GLM 4.7

    z-ai/glm-4.7-20251222

    85.7%

    View modelEvidence Evidence class unreported

  6. Qwen3.5-27B

    qwen/qwen3.5-27b-20260224

    85.5%

    View modelEvidence Evidence class unreported

  7. MiniMax M2.5

    minimax/minimax-m2.5-20260211

    85.2%

    View modelEvidence Evidence class unreported

  8. Kimi K2 Thinking

    moonshotai/kimi-k2-thinking-20251106

    84.5%

    View modelEvidence Evidence class unreported

  9. Qwen3.5-35B-A3B

    qwen/qwen3.5-35b-a3b-20260224

    84.2%

    View modelEvidence Evidence class unreported

  10. Step 3.5 Flash

    stepfun/step-3.5-flash

    83.5%

    View modelEvidence Evidence class unreported

  11. DeepSeek V3.2

    deepseek/deepseek-v3.2-20251201

    82.4%

    View modelEvidence Evidence class unreported

  12. Qwen3.5-9B

    qwen/qwen3.5-9b-20260310

    81.7%

    View modelEvidence Evidence class unreported

  13. GLM 4.7 Flash

    z-ai/glm-4.7-flash-20260119

    75.2%

    View modelEvidence Evidence class unreported

  14. R1

    deepseek/deepseek-r1

    71.5%

    View modelEvidence Evidence class unreported

  15. Mistral Small 4

    mistralai/mistral-small-2603

    71.2%

    View modelEvidence Evidence class unreported

  16. gpt-oss-20b

    openai/gpt-oss-20b

    56.8%

    View modelEvidence Evidence class unreported

  17. Llama 3.1 8B Instruct

    meta/llama-3.1-8b-instruct

    30.4%

    View modelEvidence Evidence class unreported
GPQA Diamond model results | VibeLeaderboard