Best AI models for Multimodal Document Q&A

Rankings use the same scenario weights and cost blending as the interactive leaderboard on AI Model Analyzer. Data is min-max normalised per benchmark; missing scores are skipped without penalty.

1. 1

   Gemini 2.5 Pro

   Google
   Score 85.1Q 95.9In $1.25/M
2. 2

   Claude Opus 4

   Anthropic
   Score 73.1Q 91.4In $15.00/M
3. 3

   Gemini 2.0 Flash

   Google
   Score 69.9Q 64.4In $0.10/M
4. 4

   Llama 4 Maverick

   Meta
   Score 68.9Q 67.3In $0.27/M
5. 5

   Claude Sonnet 4

   Anthropic
   Score 66.1Q 75.3In $3.00/M
6. 6

   Llama 4 Scout

   Meta
   Score 60.2Q 54.7In $0.18/M
7. 7

   Claude 3.5 Sonnet

   Anthropic
   Score 50.2Q 55.5In $3.00/M
8. 8

   GPT-4o

   OpenAI
   Score 45.5Q 48.5In $2.50/M
9. 9

   Gemini 1.5 Pro

   Google
   Score 44.7Q 44.3In $1.25/M
10. 10

    Gemini 1.5 Flash

    Google
    Score 34.3Q 18.6In $0.08/M
11. 11

    GPT-4o mini

    OpenAI
    Score 34.0Q 21.3In $0.15/M
12. 12

    GPT-5 nano

    OpenAI
    Score 20.0Q 0.0In $0.05/M
13. 13

    GPT-5 mini

    OpenAI
    Score 14.2Q 0.0In $0.25/M
14. 14

    Gemini 2.5 Flash

    Google
    Score 13.4Q 0.0In $0.30/M
15. 15

    Gemini 3 Flash

    Google
    Score 13.4Q 0.0In $0.30/M