Best value on RULER 128k

“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

   Gemini 1.5 Flash

   Google
   1116.27

   83.7 / $0.08/M
2. 2

   Gemini 2.0 Flash

   Google
   953.50

   95.3 / $0.10/M
3. 3

   GPT-4o mini

   OpenAI
   465.13

   69.8 / $0.15/M
4. 4

   Qwen3 235B

   Alibaba (Qwen)
   418.60

   83.7 / $0.20/M
5. 5

   Llama 4 Scout

   Meta
   400.50

   72.1 / $0.18/M
6. 6

   Llama 4 Maverick

   Meta
   327.30

   88.4 / $0.27/M
7. 7

   DeepSeek V3

   DeepSeek
   292.85

   79.1 / $0.27/M
8. 8

   DeepSeek R1

   DeepSeek
   152.22

   83.7 / $0.55/M
9. 9

   Claude 3.5 Haiku

   Anthropic
   87.21

   69.8 / $0.80/M
10. 10

    Gemini 2.5 Pro

    Google
    80.00

    100.0 / $1.25/M
11. 11

    Gemini 1.5 Pro

    Google
    78.14

    97.7 / $1.25/M
12. 12

    o3-mini

    OpenAI
    76.11

    83.7 / $1.10/M
13. 13

    Llama 3.3 70B Instruct

    Meta
    66.07

    58.1 / $0.88/M
14. 14

    Qwen2.5 72B Instruct

    Alibaba (Qwen)
    64.60

    58.1 / $0.90/M
15. 15

    Llama 3.1 70B Instruct

    Meta
    55.50

    48.8 / $0.88/M
16. 16

    Grok 2

    xAI
    34.88

    69.8 / $2.00/M
17. 17

    GPT-4o

    OpenAI
    33.49

    83.7 / $2.50/M
18. 18

    Claude Sonnet 4

    Anthropic
    31.01

    93.0 / $3.00/M
19. 19

    Grok 3

    xAI
    31.01

    93.0 / $3.00/M
20. 20

    Claude 3.5 Sonnet

    Anthropic
    29.46

    88.4 / $3.00/M
21. 21

    o1-mini

    OpenAI
    26.36

    79.1 / $3.00/M
22. 22

    Mistral Large 2

    Mistral
    23.25

    46.5 / $2.00/M
23. 23

    Llama 3.1 405B Instruct

    Meta
    16.61

    58.1 / $3.50/M
24. 24

    o3

    OpenAI
    9.77

    97.7 / $10.00/M
25. 25

    Claude Opus 4

    Anthropic
    6.51

    97.7 / $15.00/M

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