modelgrep

Best LLMs for Math & Science

Match · Updated July 2026

The best LLM for math and science is GPT-5.6 Sol, scoring 94.1% on GPQA Diamond (graduate-level physics, chemistry and biology). Gemini 3.1 Pro Preview (94.1%) and Kimi K3 (93.5%) round out the top three.

94.1%GPQA
58.9Intelligence
74 t/sSpeed
$5.00Input /M
1.1MContext

AI models ranked by GPQA Diamond — graduate-level physics, chemistry and biology questions that can't be answered by lookup. The best LLMs for math, quantitative reasoning and hard-science work.

  1. 1openai logo
    gpt-5.6-sol
    ReasoningToolsJSON+158.9 intel · $5.00/M · 74 t/s
    94.1%
    GPQA
  2. 2google logo
    gemini-3.1-pro-preview
    ReasoningToolsJSON+246.5 intel · $2.00/M · 122 t/s
    94.1%
    GPQA
  3. 3moonshotai logo
    kimi-k3
    ReasoningToolsJSON+157.1 intel · $3.00/M · 1.0M ctx
    93.5%
    GPQA
  4. 4openai logo
    gpt-5.5
    ReasoningToolsJSON+154.8 intel · $5.00/M · 100 t/s
    93.5%
    GPQA
  5. 5x-ai logo
    grok-4.5
    ReasoningToolsJSON+153.8 intel · $2.00/M · 500K ctx
    93.1%
    GPQA
  6. 6minimax logo
    minimax-m3
    ReasoningToolsJSON+144.4 intel · $0.300/M · 1.0M ctx
    92.9%
    GPQA
  7. 7google logo
    gemini-3.6-flash
    ReasoningToolsJSON+250.1 intel · $1.50/M · 111 t/s
    92.8%
    GPQA
  8. 8anthropic logo
    claude-fable-5
    ReasoningToolsJSON+159.9 intel · $10.00/M · 1M ctx
    92.6%
    GPQA
  9. 9openai logo
    gpt-5.6-terra
    ReasoningToolsJSON+155.0 intel · $2.50/M · 96 t/s
    92.5%
    GPQA
  10. 10qwen logo
    qwen3.7-max
    ReasoningToolsJSON46.0 intel · $1.48/M · 1M ctx
    92.3%
    GPQA
  11. 11google logo
    gemini-3.5-flash
    ReasoningToolsJSON+250.2 intel · $1.50/M · 189 t/s
    92.2%
    GPQA
  12. 12anthropic logo
    claude-opus-4.8
    ReasoningToolsJSON+155.7 intel · $5.00/M · 1M ctx
    92.0%
    GPQA
  13. 13openai logo
    gpt-5.4
    ReasoningToolsJSON+151.4 intel · $2.50/M · 79 t/s
    92.0%
    GPQA
  14. 14openai logo
    gpt-5.3-codex
    ReasoningToolsJSON+144.3 intel · $1.75/M · 400K ctx
    91.5%
    GPQA
  15. 15anthropic logo
    claude-opus-4.7
    ReasoningToolsJSON+153.5 intel · $5.00/M · 1M ctx
    91.4%
    GPQA
  16. 16openai logo
    gpt-5.6-luna
    ReasoningToolsJSON+151.2 intel · $1.00/M · 97 t/s
    91.1%
    GPQA
  17. 17anthropic logo
    claude-sonnet-5
    ReasoningToolsJSON+153.4 intel · $2.00/M · 1M ctx
    91.1%
    GPQA
  18. 18moonshotai logo
    kimi-k2.6
    ReasoningToolsJSON+144.2 intel · $0.684/M · 262K ctx
    91.1%
    GPQA
  19. 19x-ai logo
    grok-4.20
    ReasoningToolsJSON+137.0 intel · $1.25/M · 2M ctx
    91.1%
    GPQA
  20. 20openai logo
    gpt-5.2
    ReasoningToolsJSON+142.2 intel · $1.75/M · 71 t/s
    90.3%
    GPQA
  21. 21x-ai logo
    grok-4.3
    ReasoningToolsJSON+137.6 intel · $1.25/M · 18 t/s
    90.1%
    GPQA
  22. 22qwen logo
    qwen3.7-plus
    ReasoningToolsJSON+139.0 intel · $0.320/M · 1M ctx
    90.0%
    GPQA
  23. 23openai logo
    gpt-5.2-codex
    ReasoningToolsJSON+140.1 intel · $1.75/M · 400K ctx
    89.9%
    GPQA
  24. 24M
    muse-spark-1.1
    ReasoningToolsJSON+250.6 intel · $1.25/M · 1.0M ctx
    89.8%
    GPQA
  25. 25tencent logo
    hy3
    ReasoningToolsJSON41.2 intel · $0.132/M · 262K ctx
    89.7%
    GPQA

Frequently asked

What is the best LLM for math?

The best LLM for math and science is GPT-5.6 Sol, scoring 94.1% on GPQA Diamond (graduate-level physics, chemistry and biology). Gemini 3.1 Pro Preview (94.1%) and Kimi K3 (93.5%) round out the top three.

Which AI is best at math?

The best AI model for math and science is GPT-5.6 Sol, scoring 94.1% on GPQA Diamond (graduate-level physics, chemistry and biology). Gemini 3.1 Pro Preview (94.1%) and Kimi K3 (93.5%) round out the top three.

What's a good alternative to GPT-5.6 Sol?

Gemini 3.1 Pro Preview (94.1%) is the closest alternative on this metric, followed by Kimi K3 (93.5%). See the full ranking above for the tradeoffs.

By maker

All rankings