Skip to content

Value ranking

Best value on HumanEval

As of October 25, 2026, the best value on HumanEval is Qwen3 235B, at 920.89 normalized points per dollar of input price ($0.09 per million tokens, score 82.9).

164 hand-written Python programming problems scored by passing unit tests. Saturated for frontier models.

“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)
    920.89
    82.9 / $0.09/M
  2. 2
    Gemini 2.0 Flash
    Google
    680.20
    68.0 / $0.10/M
  3. 3
    Llama 3.3 70B Instruct
    Meta
    635.10
    63.5 / $0.10/M
  4. 4
    Llama 4 Scout
    Meta
    527.00
    52.7 / $0.10/M
  5. 5
    Llama 4 Maverick
    Meta
    391.57
    73.4 / $0.19/M
  6. 6
    GPT-4o mini
    OpenAI
    387.40
    58.1 / $0.15/M
  7. 7
    DeepSeek V3
    DeepSeek
    214.34
    62.2 / $0.29/M
  8. 8
    DeepSeek R1
    DeepSeek
    198.20
    99.1 / $0.50/M
  9. 9
    Qwen2.5 72B Instruct
    Alibaba (Qwen)
    153.92
    55.4 / $0.36/M
  10. 10
    Gemini 2.5 Pro
    Google
    152.80
    95.5 / $0.63/M
  11. 11
    o3-mini
    OpenAI
    78.63
    86.5 / $1.10/M
  12. 12
    Claude 3.5 Haiku
    Anthropic
    77.70
    62.2 / $0.80/M
  13. 13
    Llama 3.1 70B Instruct
    Meta
    69.82
    27.9 / $0.40/M
  14. 14
    o3
    OpenAI
    46.84
    93.7 / $2.00/M
  15. 15
    Gemini 1.5 Pro
    Google
    35.31
    44.1 / $1.25/M
  16. 16
    Grok 2
    xAI
    31.75
    63.5 / $2.00/M
  17. 17
    Claude Sonnet 4
    Anthropic
    31.08
    93.2 / $3.00/M
  18. 18
    Claude 3.5 Sonnet
    Anthropic
    29.13
    87.4 / $3.00/M
  19. 19
    Mistral Large 2
    Mistral
    29.05
    58.1 / $2.00/M
  20. 20
    GPT-4o
    OpenAI
    28.65
    71.6 / $2.50/M
  21. 21
    o1-mini
    OpenAI
    27.18
    81.5 / $3.00/M
  22. 22
    Grok 3
    xAI
    25.83
    77.5 / $3.00/M
  23. 23
    Llama 3.1 405B Instruct
    Meta
    18.92
    66.2 / $3.50/M
  24. 24
    Claude Opus 4
    Anthropic
    6.67
    100.0 / $15.00/M
  25. 25
    o1
    OpenAI
    5.44
    81.5 / $15.00/M
Open full leaderboard

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