Benchmark Leaderboard
Top models by Google-Proof Q&A score
Sorts are only valid inside this benchmark. A score gap of this benchmark does not map to the same gap on another benchmark, so validate any short-list by checking peer rows and model context/price fit.
| # | Model | Score | Version | Source |
|---|---|---|---|---|
| 1 | Claude Mythos Preview | 94.6 | diamond | https://epoch.ai/benchmarks/gpqa-diamond; https://intuitionlabs.ai/articles/gpqa-diamond-ai-benchmark |
| 2 | GPT-5.6 Sol | 94.6 | GPQA Diamond; max reasoning | https://openai.com/index/gpt-5-6/ |
| 3 | Gemini 3.1 Pro Preview | 94.3 | diamond | https://artificialanalysis.ai/leaderboards/models |
| 4 | Claude Opus 4.7 | 94.2 | diamond | https://www.anthropic.com/news/claude-opus-4-7 |
| 5 | GPT-5.5 | 93.6 | diamond | https://openai.com/index/introducing-gpt-5-5/ |
| 6 | GPT-5.5 Pro | 93.6 | GPQA Diamond | https://codingfleet.com/blog/claude-opus-4-8-vs-gpt-5-5-comparison/ |
| 7 | Claude Opus 4.8 | 93.6 | GPQA Diamond | https://www.anthropic.com/news/claude-opus-4-8 |
| 8 | Kimi K3 | 93.5 | GPQA-Diamond; max reasoning effort | https://www.kimi.com/blog/kimi-k3 |
| 9 | MiniMax M3 | 92.9 | MiniMax M3 GPQA 92 (accuracy%) | https://venturebeat.com/technology/minimax-m3-debuts-eclipsing-gpt-5-5-and-gemini-3-1-pro-on-key-benchmark-performance-for-just-5-10-of-the-cost |
| 10 | Qwen3.7-Max | 92.4 | diamond | https://www.datacamp.com/blog/qwen3-7-max |
| 11 | Gemini 3.5 Flash | 92.2 | GPQA Diamond (accuracy) | https://www.nxcode.io/resources/news/gemini-3-5-flash-complete-guide-benchmarks-pricing-api-2026 |
| 12 | GPT-5.4 | 92 | diamond | https://pricepertoken.com/leaderboards/benchmark/gpqa |
| 13 | Gemini 3 Pro | 91.9 | — | https://deepmind.google/technologies/gemini/pro/ |
| 14 | Qwen3.6-Max | 91.8 | — | https://qwenlm.github.io/blog/qwen3.6/ |
| 15 | Claude Opus 4.6 | 91.3 | diamond | https://www.anthropic.com/claude/opus |
| 16 | GLM-5.2 | 91.2 | diamond | https://huggingface.co/zai-org/GLM-5.2 |
| 17 | Kimi K2.6 | 90.5 | GPQA Diamond (accuracy) | https://huggingface.co/moonshotai/Kimi-K2.6 |
| 18 | Gemini 3 Flash | 90.4 | — | https://deepmind.google/technologies/gemini/flash/ |
| 19 | Qwen3.6-Plus | 90.4 | llm-stats shows 0 (accuracy%) | https://llm-stats.com/benchmarks/gpqa |
| 20 | DeepSeek V4 Pro | 90.1 | diamond | https://api-docs.deepseek.com/news/news260424 |
Interpretation
Trust when:
- Use top rows for shortlisting, then validate provider coverage and feature fit before selecting a model.
- Cross-benchmark consistency is a stronger signal than one-off score dominance.
Don't trust when:
- Don't compare raw score gaps across different benchmark scales as equivalent.
- Missing rows in the heatmap can hide benchmark blind spots.
- No trend data means this is a snapshot; expect drift in fresh refresh cycles.
Cross-benchmark heatmap
Compare top models across selected benchmark families to spot strengths and weak spots. On narrow screens, scroll horizontally to inspect every benchmark.
| Model | MMLU | GPQA | HumanEval | HellaSwag |
|---|---|---|---|---|
| Claude Mythos Preview | — | 94.6 | — | — |
| GPT-5.6 Sol | — | 94.6 | — | — |
| Gemini 3.1 Pro Preview | 98.0 | 94.3 | 94.0 | — |
| Claude Opus 4.7 | — | 94.2 | — | — |
| GPT-5.5 | 92.4 | 93.6 | 94.2 | — |
| GPT-5.5 Pro | — | 93.6 | — | — |