Acropolis

Vibe Code Bench

Vibe Code Bench, a component of the Vals Index.

Upstream source
CodingAgentic
Independent. Run by an operator independent of the model vendors.Vals AI benchmarks by Vals AIVersion 1.1Data as of 17 Nov 2026

No filters. Showing 38 of 38 rows.

  1. 1
    AnthropicDossier

    high · OpenHands

    Accuracy

    Fable 5.2, Accuracy: 93.8% (scale 0.0 to 100.0), 95% confidence interval 91.5 to 96.093.8%(91.5–96.0)
    Duration per test
    2,437.3 s
    Cost per test
    $22.13
  2. 2
    AnthropicDossier

    max · OpenHands

    Tied #1

    Accuracy

    Sonnet 5.5, Accuracy: 92.4% (scale 0.0 to 100.0), 95% confidence interval 89.9 to 94.992.4%(89.9–94.9)
    Duration per test
    3,672.1 s
    Cost per test
    $19.73
  3. 3
    GoogleDossier

    high · OpenHands

    Tied #1

    Accuracy

    Gemini 4 Argon, Accuracy: 91.9% (scale 0.0 to 100.0), 95% confidence interval 88.2 to 95.691.9%(88.2–95.6)
    Duration per test
    2,247.7 s
    Cost per test
    $8.11
  4. 4
    AnthropicDossier

    max · OpenHands

    Tied #1

    Accuracy

    Haiku 5.5, Accuracy: 90.4% (scale 0.0 to 100.0), 95% confidence interval 87.1 to 93.890.4%(87.1–93.8)
    Duration per test
    2,075.7 s
    Cost per test
    $6.07
  5. 5
    AnthropicDossier

    max · OpenHands

    Tied #1

    Accuracy

    Opus 5.5, Accuracy: 90.3% (scale 0.0 to 100.0), 95% confidence interval 87.3 to 93.390.3%(87.3–93.3)
    Duration per test
    5,766.5 s
    Cost per test
    $35.77
  6. 6
    OpenAIDossier

    max · OpenHands

    Tied #1

    Accuracy

    GPT-6 Astra, Accuracy: 89.6% (scale 0.0 to 100.0), 95% confidence interval 85.3 to 93.989.6%(85.3–93.9)
    Duration per test
    2,344.1 s
    Cost per test
    $38.51
  7. 7
    GoogleDossier

    high · OpenHands

    Tied #1

    Accuracy

    Gemini 4 Carbon, Accuracy: 89.1% (scale 0.0 to 100.0), 95% confidence interval 84.6 to 93.689.1%(84.6–93.6)
    Duration per test
    2,172 s
    Cost per test
    $8.25
  8. 8
    OpenAIDossier

    max · OpenHands

    Accuracy

    GPT-6.1 Sol, Accuracy: 88.9% (scale 0.0 to 100.0), 95% confidence interval 85.2 to 92.688.9%(85.2–92.6)
    Duration per test
    4,489.8 s
    Cost per test
    $6.23
  9. 9
    GoogleDossier

    xhigh · OpenHands

    Accuracy

    Gemini 4 Hydrogen, Accuracy: 87.8% (scale 0.0 to 100.0), 95% confidence interval 83.3 to 92.387.8%(83.3–92.3)
    Duration per test
    864.6 s
    Cost per test
    $4.55
  10. 10
    DeepSeekDossier

    max · OpenHands

    Accuracy

    DeepSeek V4.1 Flash 1017, Accuracy: 87.7% (scale 0.0 to 100.0), 95% confidence interval 82.3 to 93.187.7%(82.3–93.1)
    Duration per test
    973.8 s
    Cost per test
    $0.58
  11. 11
    AlibabaDossier

    xhigh · OpenHands

    Accuracy

    Qwen4 Max, Accuracy: 86.7% (scale 0.0 to 100.0), 95% confidence interval 82.2 to 91.186.7%(82.2–91.1)
    Duration per test
    4,172.5 s
    Cost per test
    $41.40
  12. 12

    xhigh · OpenHands

    Accuracy

    Grok 4.7, Accuracy: 86.2% (scale 0.0 to 100.0), 95% confidence interval 81.9 to 90.486.2%(81.9–90.4)
    Duration per test
    2,147.1 s
    Cost per test
    $15.83
  13. 13

    max · OpenHands

    Accuracy

    Muse Wonder Full 1.0, Accuracy: 86% (scale 0.0 to 100.0), 95% confidence interval 82.2 to 89.886%(82.2–89.8)
    Duration per test
    648.3 s
    Cost per test
    $2.29
  14. 14
    XiaomiDossier

    OpenHands

    Accuracy

    MiMo-V2.6-Pro, Accuracy: 85.2% (scale 0.0 to 100.0), 95% confidence interval 78.6 to 91.985.2%(78.6–91.9)
    Duration per test
    3,646.6 s
    Cost per test
    $1.04
  15. 15
    Moonshot AIDossier

    max · OpenHands

    Accuracy

    Kimi K3, Accuracy: 85% (scale 0.0 to 100.0), 95% confidence interval 79.6 to 90.485%(79.6–90.4)
    Duration per test
    999 s
    Cost per test
    $10.01
  16. 16
    Moonshot AIDossier

    max · OpenHands

    Accuracy

    Kimi K3.1 Pro, Accuracy: 84.9% (scale 0.0 to 100.0), 95% confidence interval 79.1 to 90.784.9%(79.1–90.7)
    Duration per test
    3,028.9 s
    Cost per test
    $7.49
  17. 17
    OpenAIDossier

    max · OpenHands

    Accuracy

    GPT-6.1 Terra, Accuracy: 84% (scale 0.0 to 100.0), 95% confidence interval 78.4 to 89.684%(78.4–89.6)
    Duration per test
    2,250 s
    Cost per test
    $6.00
  18. 18

    max · OpenHands

    Accuracy

    Muse Ember, Accuracy: 83.7% (scale 0.0 to 100.0), 95% confidence interval 78.3 to 89.183.7%(78.3–89.1)
    Duration per test
    167.9 s
    Cost per test
    $0.054
  19. 19

    max · OpenHands

    Accuracy

    Muse Wonder 1.0, Accuracy: 83.1% (scale 0.0 to 100.0), 95% confidence interval 78.2 to 88.083.1%(78.2–88.0)
    Duration per test
    350.1 s
    Cost per test
    $1.80
  20. 20
    DeepSeekDossier

    max · OpenHands

    Accuracy

    DeepSeek V4 Pro 0813, Accuracy: 82.3% (scale 0.0 to 100.0), 95% confidence interval 76.1 to 88.582.3%(76.1–88.5)
    Duration per test
    3,982.6 s
    Cost per test
    $0.36
  21. 21
    OpenAIDossier

    max · OpenHands

    Accuracy

    GPT-6 Luna, Accuracy: 81.6% (scale 0.0 to 100.0), 95% confidence interval 75.0 to 88.381.6%(75.0–88.3)
    Duration per test
    2,219.5 s
    Cost per test
    $1.35
  22. 22
    MistralDossier

    high · OpenHands

    Accuracy

    Mistral Large 4 OSS, Accuracy: 81.6% (scale 0.0 to 100.0), 95% confidence interval 75.1 to 88.081.6%(75.1–88.0)
    Duration per test
    3,012.6 s
    Cost per test
    $11.72
  23. 23
    Thinking MachinesDossier

    OpenHands

    Accuracy

    Inkling 2 Small, Accuracy: 81.5% (scale 0.0 to 100.0), 95% confidence interval 74.8 to 88.381.5%(74.8–88.3)
    Duration per test
    1,537.4 s
    Cost per test
    $1.90
  24. 24
    TencentDossier

    OpenHands

    Accuracy

    Hy4, Accuracy: 81% (scale 0.0 to 100.0), 95% confidence interval 74.1 to 88.081%(74.1–88.0)
    Duration per test
    2,281.5 s
    Cost per test
    $2.39
  25. 25
    XiaomiDossier

    OpenHands

    Accuracy

    MiMo-V2.6-Flash, Accuracy: 79% (scale 0.0 to 100.0), 95% confidence interval 71.0 to 86.979%(71.0–86.9)
    Duration per test
    3,013 s
    Cost per test
    $0.56
  26. 26
    GoogleDossier

    high · OpenHands

    Accuracy

    Gemini 3.8 Flash, Accuracy: 78.7% (scale 0.0 to 100.0), 95% confidence interval 71.0 to 86.378.7%(71.0–86.3)
    Duration per test
    519.1 s
    Cost per test
    $6.87
  27. 27

    max · OpenHands

    Accuracy

    GLM-5.3, Accuracy: 78.1% (scale 0.0 to 100.0), 95% confidence interval 70.4 to 85.878.1%(70.4–85.8)
    Duration per test
    3,847.2 s
    Cost per test
    $12.45
  28. 28
    TencentDossier

    OpenHands

    Accuracy

    Hy4 Preview, Accuracy: 77.5% (scale 0.0 to 100.0), 95% confidence interval 69.6 to 85.477.5%(69.6–85.4)
    Duration per test
    2,493.1 s
    Cost per test
    $2.18
  29. 29
    AlibabaDossier

    OpenHands

    Accuracy

    Qwen3.8 Max Internal, Accuracy: 77.1% (scale 0.0 to 100.0), 95% confidence interval 69.8 to 84.477.1%(69.8–84.4)
    Duration per test
    5,079.4 s
    Cost per test
    $11.67
  30. 30

    xhigh · OpenHands

    Accuracy

    Muse Spark Open, Accuracy: 76.2% (scale 0.0 to 100.0), 95% confidence interval 69.7 to 82.776.2%(69.7–82.7)
    Duration per test
    1,869.2 s
    Cost per test
    $1.66
  31. 31
    MiniMaxDossier

    max · OpenHands

    Accuracy

    MiniMax M3.1 Flash, Accuracy: 73.9% (scale 0.0 to 100.0), 95% confidence interval 65.6 to 82.373.9%(65.6–82.3)
    Duration per test
    1,263.9 s
    Cost per test
    $1.15
  32. 32
    StepFunDossier

    OpenHands

    Accuracy

    Step 5 Preview, Accuracy: 70.4% (scale 0.0 to 100.0), 95% confidence interval 62.0 to 78.870.4%(62.0–78.8)
    Duration per test
    5,466.4 s
    Cost per test
    $5.86
  33. 33
    ReflectionDossier

    OpenHands

    Accuracy

    Beam 1.0, Accuracy: 66.4% (scale 0.0 to 100.0), 95% confidence interval 58.5 to 74.466.4%(58.5–74.4)
    Duration per test
    145.3 s
    Cost per test
    Not tested
  34. 34
    InceptionDossier

    high · OpenHands

    Accuracy

    Mercury 3, Accuracy: 60.9% (scale 0.0 to 100.0), 95% confidence interval 50.4 to 71.560.9%(50.4–71.5)
    Duration per test
    566.6 s
    Cost per test
    $0.63
  35. 35

    max · OpenHands

    Accuracy

    GLM-5.3 Flash, Accuracy: 30.8% (scale 0.0 to 100.0), 95% confidence interval 20.4 to 41.130.8%(20.4–41.1)
    Duration per test
    5,027.3 s
    Cost per test
    $4.70
  36. 36
    Thinking MachinesDossier

    OpenHands

    Accuracy

    Inkling, Accuracy: 19.2% (scale 0.0 to 100.0), 95% confidence interval 12.0 to 26.419.2%(12.0–26.4)
    Duration per test
    1,105.5 s
    Cost per test
    $1.61
  37. 37
    Thinking MachinesDossier

    OpenHands

    Accuracy

    Inkling Small, Accuracy: 19.1% (scale 0.0 to 100.0), 95% confidence interval 12.4 to 25.719.1%(12.4–25.7)
    Duration per test
    2,949.5 s
    Cost per test
    $0.43
  38. 38
    NVIDIADossier

    high · OpenHands

    Accuracy

    Nemotron 3.5 Super VL, Accuracy: 3.7% (scale 0.0 to 100.0), 95% confidence interval 1.3 to 10.03.7%(1.3–10.0)
    Duration per test
    262.9 s
    Cost per test
    $0.54