Best value on AIME 2024

“Value” is normalized benchmark score (0–100 for this leaderboard cohort) divided by input price per million tokens. Higher means more capability per dollar on this axis only — always sanity-check latency, context length, and your real workload.

1. 1

   Qwen3 235B

   Alibaba (Qwen)
   433.40

   86.7 / $0.20/M
2. 2

   Gemini 2.0 Flash

   Google
   308.10

   30.8 / $0.10/M
3. 3

   Llama 4 Scout

   Meta
   210.67

   37.9 / $0.18/M
4. 4

   Llama 4 Maverick

   Meta
   170.56

   46.0 / $0.27/M
5. 5

   DeepSeek R1

   DeepSeek
   154.11

   84.8 / $0.55/M
6. 6

   DeepSeek V3

   DeepSeek
   144.22

   38.9 / $0.27/M
7. 7

   o3-mini

   OpenAI
   84.75

   93.2 / $1.10/M
8. 8

   Gemini 2.5 Pro

   Google
   78.82

   98.5 / $1.25/M
9. 9

   Gemini 1.5 Flash

   Google
   34.67

   2.6 / $0.08/M
10. 10

    Grok 3

    xAI
    33.33

    100.0 / $3.00/M
11. 11

    Llama 3.3 70B Instruct

    Meta
    32.45

    28.6 / $0.88/M
12. 12

    GPT-4o mini

    OpenAI
    28.60

    4.3 / $0.15/M
13. 13

    Qwen2.5 72B Instruct

    Alibaba (Qwen)
    23.32

    21.0 / $0.90/M
14. 14

    o1-mini

    OpenAI
    19.56

    58.7 / $3.00/M
15. 15

    Llama 3.1 70B Instruct

    Meta
    15.39

    13.5 / $0.88/M
16. 16

    Claude Sonnet 4

    Anthropic
    12.04

    36.1 / $3.00/M
17. 17

    Gemini 1.5 Pro

    Google
    11.10

    13.9 / $1.25/M
18. 18

    o3

    OpenAI
    9.50

    95.0 / $10.00/M
19. 19

    Mistral Large 2

    Mistral
    6.94

    13.9 / $2.00/M
20. 20

    Grok 2

    xAI
    6.77

    13.5 / $2.00/M
21. 21

    Llama 3.1 405B Instruct

    Meta
    5.90

    20.6 / $3.50/M
22. 22

    o1

    OpenAI
    5.61

    84.1 / $15.00/M
23. 23

    Claude Opus 4

    Anthropic
    5.33

    79.9 / $15.00/M
24. 24

    Claude 3.5 Sonnet

    Anthropic
    4.25

    12.8 / $3.00/M
25. 25

    GPT-4o

    OpenAI
    3.93

    9.8 / $2.50/M

AI Model Analyzer does not recommend specific vendors; rankings are derived from public data only.