Skip to content

Value ranking

Best value on LiveCodeBench

As of October 25, 2026, the best value on LiveCodeBench is Qwen3 235B, at 962.67 normalized points per dollar of input price ($0.09 per million tokens, score 86.6).

Continuously refreshed coding benchmark drawing from LeetCode, AtCoder, and Codeforces; reduces benchmark contamination.

“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)
    962.67
    86.6 / $0.09/M
  2. 2
    Gemini 2.5 Flash
    Google
    523.87
    78.6 / $0.15/M
  3. 3
    DeepSeek R1
    DeepSeek
    188.66
    94.3 / $0.50/M
  4. 4
    Gemini 2.5 Pro
    Google
    150.62
    94.1 / $0.63/M
  5. 5
    DeepSeek V3
    DeepSeek
    93.48
    27.1 / $0.29/M
  6. 6
    o4-mini
    OpenAI
    90.91
    100.0 / $1.10/M
  7. 7
    o3-mini
    OpenAI
    74.10
    81.5 / $1.10/M
  8. 8
    o3
    OpenAI
    47.53
    95.1 / $2.00/M
  9. 9
    Claude Sonnet 4 (Thinking)
    Anthropic
    21.25
    63.8 / $3.00/M
  10. 10
    Claude Sonnet 4
    Anthropic
    15.39
    46.2 / $3.00/M
  11. 11
    Claude 3.5 Sonnet
    Anthropic
    8.51
    25.5 / $3.00/M
  12. 12
    Claude Opus 4 (Thinking)
    Anthropic
    4.49
    67.4 / $15.00/M
  13. 13
    Claude Opus 4
    Anthropic
    3.46
    51.9 / $15.00/M
  14. 14
    GPT-4o
    OpenAI
    2.15
    5.4 / $2.50/M
  15. 15
    GPT-4 Turbo
    OpenAI
    0.35
    3.5 / $10.00/M
  16. 16
    GPT-4o mini
    OpenAI
    0.00
    0.0 / $0.15/M
Open full leaderboard

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