Acropolis

LMArena Text (style control)

Arena human-vote leaderboard: rating with confidence interval, vote count and rank. One row per model and category.

General text
Independent. Run by an operator independent of the model vendors.LMArena by ArenaVersion 1Data as of 17 Nov 2026

No filters. Showing 421 of 421 rows.

  1. 1
    GoogleDossier

    high

    Arena rating

    Gemini 4 Carbon, Arena rating: 1,526 (scale 900.0 to 1526.2), 95% confidence interval 1516.6 to 1535.81,526(1516.6–1535.8)
  2. 2
    GoogleDossier

    high

    Tied #1

    Arena rating

    Gemini 4 Argon, Arena rating: 1,522 (scale 900.0 to 1526.2), 95% confidence interval 1516.3 to 1527.11,522(1516.3–1527.1)
  3. 3

    max

    Arena rating

    Muse Wonder 1.0, Arena rating: 1,511 (scale 900.0 to 1526.2), 95% confidence interval 1504.7 to 1517.11,511(1504.7–1517.1)
  4. 4
    AnthropicDossier

    high

    Arena rating

    Opus 5.5, Arena rating: 1,507 (scale 900.0 to 1526.2), 95% confidence interval 1500.8 to 1512.71,507(1500.8–1512.7)
  5. 5
    AnthropicDossier

    high

    Arena rating

    Fable 5, Arena rating: 1,504 (scale 900.0 to 1526.2), 95% confidence interval 1499.8 to 1508.21,504(1499.8–1508.2)
  6. 6
    AnthropicDossier

    high

    Arena rating

    Opus 4.6, Arena rating: 1,503 (scale 900.0 to 1526.2), 95% confidence interval 1499.6 to 1506.11,503(1499.6–1506.1)
  7. 7
    AnthropicDossier

    max

    Arena rating

    Fable 5.1, Arena rating: 1,502 (scale 900.0 to 1526.2), 95% confidence interval 1496.8 to 1507.81,502(1496.8–1507.8)
  8. 8
    AnthropicDossier

    high

    Arena rating

    Opus 4.7, Arena rating: 1,501 (scale 900.0 to 1526.2), 95% confidence interval 1497.4 to 1504.91,501(1497.4–1504.9)
  9. 9
    GoogleDossier

    high

    Arena rating

    Gemini 3.8 Flash, Arena rating: 1,497 (scale 900.0 to 1526.2), 95% confidence interval 1492.5 to 1500.91,497(1492.5–1500.9)
  10. 10

    max

    Arena rating

    Muse Spark 1.3, Arena rating: 1,497 (scale 900.0 to 1526.2), 95% confidence interval 1491.5 to 1501.51,497(1491.5–1501.5)
  11. 11
    AnthropicDossier

    Arena rating

    Opus 4.6, Arena rating: 1,496 (scale 900.0 to 1526.2), 95% confidence interval 1493.2 to 1499.61,496(1493.2–1499.6)
  12. 12
    AnthropicDossier

    Arena rating

    Opus 4.7, Arena rating: 1,494 (scale 900.0 to 1526.2), 95% confidence interval 1490.1 to 1497.61,494(1490.1–1497.6)
  13. 13

    xhigh

    Arena rating

    Muse Spark 1.2, Arena rating: 1,493 (scale 900.0 to 1526.2), 95% confidence interval 1487.1 to 1498.71,493(1487.1–1498.7)
  14. 14

    Arena rating

    Muse Spark 1.1, Arena rating: 1,491 (scale 900.0 to 1526.2), 95% confidence interval 1486.7 to 1495.01,491(1486.7–1495.0)
  15. 15
    GoogleDossier

    high

    Arena rating

    Gemini 4 Hydrogen, Arena rating: 1,490 (scale 900.0 to 1526.2), 95% confidence interval 1481.4 to 1499.41,490(1481.4–1499.4)
  16. 16
    AnthropicDossier

    high

    Arena rating

    Opus 5, Arena rating: 1,490 (scale 900.0 to 1526.2), 95% confidence interval 1486.1 to 1493.31,490(1486.1–1493.3)
  17. 17
    GoogleDossier

    high

    Arena rating

    Gemini 3.7 Flash, Arena rating: 1,489 (scale 900.0 to 1526.2), 95% confidence interval 1484.6 to 1494.01,489(1484.6–1494.0)
  18. 18

    Arena rating

    Muse Spark, Arena rating: 1,489 (scale 900.0 to 1526.2), 95% confidence interval 1483.4 to 1495.01,489(1483.4–1495.0)
  19. 19
    AnthropicDossier

    max

    Arena rating

    Opus 5, Arena rating: 1,489 (scale 900.0 to 1526.2), 95% confidence interval 1484.2 to 1493.71,489(1484.2–1493.7)
  20. 20
    Moonshot AIDossier

    max

    Arena rating

    Kimi K3, Arena rating: 1,487 (scale 900.0 to 1526.2), 95% confidence interval 1483.2 to 1491.71,487(1483.2–1491.7)
  21. 21
    GoogleDossier

    Arena rating

    Gemini 3.1 Pro Preview, Arena rating: 1,487 (scale 900.0 to 1526.2), 95% confidence interval 1483.9 to 1489.71,487(1483.9–1489.7)
  22. 22
    GoogleDossier

    Arena rating

    Gemini 3 Pro Preview, Arena rating: 1,486 (scale 900.0 to 1526.2), 95% confidence interval 1481.8 to 1489.41,486(1481.8–1489.4)
  23. 23
    OpenAIDossier

    xhigh

    Arena rating

    GPT-5.6 Sol, Arena rating: 1,485 (scale 900.0 to 1526.2), 95% confidence interval 1480.9 to 1489.41,485(1480.9–1489.4)
  24. 24
    Moonshot AIDossier

    max

    Arena rating

    Kimi K3.1 Pro, Arena rating: 1,485 (scale 900.0 to 1526.2), 95% confidence interval 1474.7 to 1494.71,485(1474.7–1494.7)
  25. 25
    AlibabaDossier

    Arena rating

    Qwen3.8-Max, Arena rating: 1,484 (scale 900.0 to 1526.2), 95% confidence interval 1479.5 to 1488.71,484(1479.5–1488.7)
  26. 26
    GoogleDossier

    high

    Arena rating

    Gemini 3.6 Flash, Arena rating: 1,483 (scale 900.0 to 1526.2), 95% confidence interval 1478.2 to 1487.11,483(1478.2–1487.1)
  27. 27
    AnthropicDossier

    high

    Arena rating

    Opus 4.8, Arena rating: 1,482 (scale 900.0 to 1526.2), 95% confidence interval 1478.0 to 1485.81,482(1478.0–1485.8)
  28. 28

    max

    Arena rating

    GLM-5.3, Arena rating: 1,482 (scale 900.0 to 1526.2), 95% confidence interval 1476.9 to 1486.31,482(1476.9–1486.3)
  29. 29
    OpenAIDossier

    high

    Arena rating

    GPT-5.5, Arena rating: 1,481 (scale 900.0 to 1526.2), 95% confidence interval 1477.5 to 1485.01,481(1477.5–1485.0)
  30. 30
    OpenAIDossier

    max

    Arena rating

    GPT-6.1 Sol, Arena rating: 1,481 (scale 900.0 to 1526.2), 95% confidence interval 1474.6 to 1487.01,481(1474.6–1487.0)
  31. 31
    XiaomiDossier

    Arena rating

    MiMo-V2.6-Pro, Arena rating: 1,480 (scale 900.0 to 1526.2), 95% confidence interval 1470.4 to 1489.41,480(1470.4–1489.4)
  32. 32
    GoogleDossier

    high

    Arena rating

    Gemini 3.5 Flash, Arena rating: 1,478 (scale 900.0 to 1526.2), 95% confidence interval 1474.5 to 1482.31,478(1474.5–1482.3)
  33. 33
    OpenAIDossier

    Arena rating

    GPT-5.5, Arena rating: 1,476 (scale 900.0 to 1526.2), 95% confidence interval 1472.4 to 1479.81,476(1472.4–1479.8)
  34. 34
    GoogleDossier

    medium

    Arena rating

    Gemini 3.5 Flash, Arena rating: 1,476 (scale 900.0 to 1526.2), 95% confidence interval 1471.9 to 1480.21,476(1471.9–1480.2)
  35. 35
    OpenAIDossier

    Arena rating

    GPT-5.2 Chat, Arena rating: 1,476 (scale 900.0 to 1526.2), 95% confidence interval 1471.5 to 1479.71,476(1471.5–1479.7)
  36. 36

    Arena rating

    GLM-5.3 Flash, Arena rating: 1,475 (scale 900.0 to 1526.2), 95% confidence interval 1470.8 to 1479.71,475(1470.8–1479.7)
  37. 37
    OpenAIDossier

    high

    Arena rating

    GPT-5.4, Arena rating: 1,475 (scale 900.0 to 1526.2), 95% confidence interval 1471.3 to 1478.91,475(1471.3–1478.9)
  38. 38
    DeepSeekDossier

    max

    Arena rating

    DeepSeek V4.1 Flash, Arena rating: 1,475 (scale 900.0 to 1526.2), 95% confidence interval 1469.0 to 1481.21,475(1469.0–1481.2)
  39. 39

    max

    Arena rating

    GLM-5.2, Arena rating: 1,475 (scale 900.0 to 1526.2), 95% confidence interval 1470.9 to 1479.11,475(1470.9–1479.1)
  40. 40

    Arena rating

    Grok 4.20 Beta 1, Arena rating: 1,475 (scale 900.0 to 1526.2), 95% confidence interval 1470.0 to 1479.21,475(1470.0–1479.2)
  41. 41
    AnthropicDossier

    Arena rating

    Opus 4.8, Arena rating: 1,474 (scale 900.0 to 1526.2), 95% confidence interval 1470.6 to 1478.31,474(1470.6–1478.3)
  42. 42
    OpenAIDossier

    max

    Arena rating

    GPT-6 Astra, Arena rating: 1,474 (scale 900.0 to 1526.2), 95% confidence interval 1468.4 to 1479.41,474(1468.4–1479.4)
  43. 43
    AlibabaDossier

    Arena rating

    Qwen3.7 Max Preview, Arena rating: 1,474 (scale 900.0 to 1526.2), 95% confidence interval 1464.0 to 1483.61,474(1464.0–1483.6)
  44. 44
    AnthropicDossier

    high

    Arena rating

    Opus 4.5, Arena rating: 1,474 (scale 900.0 to 1526.2), 95% confidence interval 1469.8 to 1477.51,474(1469.8–1477.5)
  45. 45
    OpenAIDossier

    Arena rating

    GPT-5.5 Instant, Arena rating: 1,473 (scale 900.0 to 1526.2), 95% confidence interval 1468.2 to 1478.21,473(1468.2–1478.2)
  46. 46
    GoogleDossier

    Arena rating

    Gemini 3 Flash Preview, Arena rating: 1,473 (scale 900.0 to 1526.2), 95% confidence interval 1468.7 to 1477.41,473(1468.7–1477.4)
  47. 47
    AnthropicDossier

    xhigh

    Arena rating

    Sonnet 5.5, Arena rating: 1,473 (scale 900.0 to 1526.2), 95% confidence interval 1466.9 to 1478.91,473(1466.9–1478.9)
  48. 48
    AnthropicDossier

    Arena rating

    Sonnet 4.6, Arena rating: 1,472 (scale 900.0 to 1526.2), 95% confidence interval 1468.6 to 1475.61,472(1468.6–1475.6)
  49. 49

    thinking

    Arena rating

    Grok 4.20 0309, Arena rating: 1,472 (scale 900.0 to 1526.2), 95% confidence interval 1468.2 to 1475.61,472(1468.2–1475.6)
  50. 50

    Arena rating

    Grok 4.20 Multi-Agent Beta (0309), Arena rating: 1,471 (scale 900.0 to 1526.2), 95% confidence interval 1467.4 to 1474.91,471(1467.4–1474.9)
  51. 51
    AnthropicDossier

    Arena rating

    Opus 4.5, Arena rating: 1,470 (scale 900.0 to 1526.2), 95% confidence interval 1466.6 to 1472.91,470(1466.6–1472.9)
  52. 52
    OpenAIDossier

    xhigh

    Arena rating

    GPT-5.6 Terra, Arena rating: 1,468 (scale 900.0 to 1526.2), 95% confidence interval 1463.8 to 1472.11,468(1463.8–1472.1)
  53. 53
    BaiduDossier

    Arena rating

    ERNIE 5.1, Arena rating: 1,467 (scale 900.0 to 1526.2), 95% confidence interval 1462.5 to 1471.61,467(1462.5–1471.6)
  54. 54
    XiaomiDossier

    Arena rating

    MiMo-V2.5-Pro, Arena rating: 1,466 (scale 900.0 to 1526.2), 95% confidence interval 1462.9 to 1469.91,466(1462.9–1469.9)
  55. 55

    Arena rating

    Grok 4.5, Arena rating: 1,466 (scale 900.0 to 1526.2), 95% confidence interval 1461.4 to 1470.11,466(1461.4–1470.1)
  56. 56

    thinking

    Arena rating

    Grok 4.1, Arena rating: 1,465 (scale 900.0 to 1526.2), 95% confidence interval 1462.1 to 1468.51,465(1462.1–1468.5)
  57. 57
    OpenAIDossier

    Arena rating

    GPT-5.4, Arena rating: 1,465 (scale 900.0 to 1526.2), 95% confidence interval 1461.4 to 1468.81,465(1461.4–1468.8)
  58. 58
    AlibabaDossier

    Arena rating

    Qwen3.5 Max Preview, Arena rating: 1,465 (scale 900.0 to 1526.2), 95% confidence interval 1460.1 to 1469.91,465(1460.1–1469.9)
  59. 59

    Arena rating

    GLM-5.1, Arena rating: 1,465 (scale 900.0 to 1526.2), 95% confidence interval 1461.4 to 1468.51,465(1461.4–1468.5)
  60. 60
    DeepSeekDossier

    high

    Arena rating

    DeepSeek V4 Pro 0813, Arena rating: 1,464 (scale 900.0 to 1526.2), 95% confidence interval 1457.2 to 1470.61,464(1457.2–1470.6)
  61. 61
    Moonshot AIDossier

    Arena rating

    Kimi K2.6, Arena rating: 1,461 (scale 900.0 to 1526.2), 95% confidence interval 1456.4 to 1465.31,461(1456.4–1465.3)
  62. 62
    AnthropicDossier

    high

    Arena rating

    Sonnet 5, Arena rating: 1,461 (scale 900.0 to 1526.2), 95% confidence interval 1456.8 to 1464.91,461(1456.8–1464.9)
  63. 63
    AlibabaDossier

    Arena rating

    Qwen3.6 Max Preview, Arena rating: 1,460 (scale 900.0 to 1526.2), 95% confidence interval 1451.5 to 1468.01,460(1451.5–1468.0)
  64. 64

    Arena rating

    Grok 4.1, Arena rating: 1,459 (scale 900.0 to 1526.2), 95% confidence interval 1456.1 to 1462.41,459(1456.1–1462.4)
  65. 65
    GoogleDossier

    minimal

    Arena rating

    Gemini 3 Flash Preview, Arena rating: 1,458 (scale 900.0 to 1526.2), 95% confidence interval 1455.2 to 1461.21,458(1455.2–1461.2)
  66. 66
    DeepSeekDossier

    Arena rating

    DeepSeek V4-Pro, Arena rating: 1,458 (scale 900.0 to 1526.2), 95% confidence interval 1454.0 to 1461.81,458(1454.0–1461.8)
  67. 67

    Arena rating

    GLM-5, Arena rating: 1,457 (scale 900.0 to 1526.2), 95% confidence interval 1453.1 to 1461.61,457(1453.1–1461.6)
  68. 68
    OpenAIDossier

    max

    Arena rating

    GPT-6 Sol, Arena rating: 1,457 (scale 900.0 to 1526.2), 95% confidence interval 1451.7 to 1462.71,457(1451.7–1462.7)
  69. 69

    high

    Arena rating

    Grok 4.6, Arena rating: 1,457 (scale 900.0 to 1526.2), 95% confidence interval 1452.0 to 1461.31,457(1452.0–1461.3)
  70. 70
    ByteDanceDossier

    Arena rating

    Dola Seed 2.0 Pro, Arena rating: 1,456 (scale 900.0 to 1526.2), 95% confidence interval 1453.1 to 1459.91,456(1453.1–1459.9)
  71. 71
    AnthropicDossier

    high

    Arena rating

    Sonnet 4.5, Arena rating: 1,456 (scale 900.0 to 1526.2), 95% confidence interval 1453.4 to 1458.91,456(1453.4–1458.9)
  72. 72
    OpenAIDossier

    high

    Arena rating

    GPT-5.1, Arena rating: 1,455 (scale 900.0 to 1526.2), 95% confidence interval 1451.7 to 1459.11,455(1451.7–1459.1)
  73. 73
    AlibabaDossier

    Arena rating

    Qwen3.7 Plus, Arena rating: 1,455 (scale 900.0 to 1526.2), 95% confidence interval 1451.0 to 1459.71,455(1451.0–1459.7)
  74. 74
    AnthropicDossier

    Arena rating

    Sonnet 4.5, Arena rating: 1,455 (scale 900.0 to 1526.2), 95% confidence interval 1452.2 to 1458.01,455(1452.2–1458.0)
  75. 75
    DeepSeekDossier

    high

    Arena rating

    DeepSeek V4 Pro Preview, Arena rating: 1,455 (scale 900.0 to 1526.2), 95% confidence interval 1451.0 to 1458.91,455(1451.0–1458.9)
  76. 76
    GoogleDossier

    Arena rating

    Gemini 3.5 Flash-Lite, Arena rating: 1,455 (scale 900.0 to 1526.2), 95% confidence interval 1450.4 to 1459.41,455(1450.4–1459.4)
  77. 77
    TencentDossier

    Arena rating

    Hy3, Arena rating: 1,454 (scale 900.0 to 1526.2), 95% confidence interval 1448.3 to 1459.71,454(1448.3–1459.7)
  78. 78
    OpenAIDossier

    xhigh

    Arena rating

    GPT-5.6 Luna, Arena rating: 1,453 (scale 900.0 to 1526.2), 95% confidence interval 1448.9 to 1457.51,453(1448.9–1457.5)
  79. 79
    GoogleDossier

    Arena rating

    Gemma 4 31B, Arena rating: 1,452 (scale 900.0 to 1526.2), 95% confidence interval 1444.4 to 1459.31,452(1444.4–1459.3)
  80. 80
    OpenAIDossier

    Arena rating

    GPT-5.3 Chat, Arena rating: 1,450 (scale 900.0 to 1526.2), 95% confidence interval 1446.0 to 1454.51,450(1446.0–1454.5)
  81. 81
    AnthropicDossier

    thinking · 16,000 budget

    Arena rating

    Opus 4.1, Arena rating: 1,450 (scale 900.0 to 1526.2), 95% confidence interval 1446.4 to 1453.41,450(1446.4–1453.4)
  82. 82
    Moonshot AIDossier

    thinking

    Arena rating

    Kimi K2.5, Arena rating: 1,450 (scale 900.0 to 1526.2), 95% confidence interval 1446.2 to 1452.81,450(1446.2–1452.8)
  83. 83
    BaiduDossier

    Arena rating

    ERNIE 5.0 Preview (1203), Arena rating: 1,449 (scale 900.0 to 1526.2), 95% confidence interval 1442.3 to 1455.31,449(1442.3–1455.3)
  84. 84
    StepFunDossier

    Arena rating

    Step 5 Preview, Arena rating: 1,449 (scale 900.0 to 1526.2), 95% confidence interval 1442.3 to 1454.81,449(1442.3–1454.8)
  85. 85
    XiaomiDossier

    Arena rating

    MiMo-V2.6-Flash, Arena rating: 1,448 (scale 900.0 to 1526.2), 95% confidence interval 1443.2 to 1453.61,448(1443.2–1453.6)
  86. 86
    XiaomiDossier

    Arena rating

    MiMo-V2-Pro, Arena rating: 1,448 (scale 900.0 to 1526.2), 95% confidence interval 1443.4 to 1452.71,448(1443.4–1452.7)
  87. 87
    AnthropicDossier

    Arena rating

    Opus 4.1, Arena rating: 1,448 (scale 900.0 to 1526.2), 95% confidence interval 1445.0 to 1451.01,448(1445.0–1451.0)
  88. 88
    OpenAIDossier

    high

    Arena rating

    GPT-5.4 mini, Arena rating: 1,447 (scale 900.0 to 1526.2), 95% confidence interval 1443.6 to 1451.11,447(1443.6–1451.1)
  89. 89
    BaiduDossier

    Arena rating

    ERNIE 5.0 (0110), Arena rating: 1,446 (scale 900.0 to 1526.2), 95% confidence interval 1441.9 to 1449.71,446(1441.9–1449.7)
  90. 90
    GoogleDossier

    Arena rating

    Gemini 2.5 Pro, Arena rating: 1,445 (scale 900.0 to 1526.2), 95% confidence interval 1443.0 to 1447.91,445(1443.0–1447.9)
  91. 91
    OpenAIDossier

    Arena rating

    GPT-4.5 Preview, Arena rating: 1,445 (scale 900.0 to 1526.2), 95% confidence interval 1439.0 to 1450.41,445(1439.0–1450.4)
  92. 92

    xhigh

    Arena rating

    Grok 4.7, Arena rating: 1,443 (scale 900.0 to 1526.2), 95% confidence interval 1438.0 to 1448.61,443(1438.0–1448.6)
  93. 93
    AlibabaDossier

    Arena rating

    Qwen3.6 Plus, Arena rating: 1,443 (scale 900.0 to 1526.2), 95% confidence interval 1438.6 to 1446.91,443(1438.6–1446.9)
  94. 94
    OpenAIDossier

    Arena rating

    ChatGPT-4o Latest, Arena rating: 1,443 (scale 900.0 to 1526.2), 95% confidence interval 1439.9 to 1445.51,443(1439.9–1445.5)
  95. 95
    Thinking MachinesDossier

    Arena rating

    Inkling, Arena rating: 1,442 (scale 900.0 to 1526.2), 95% confidence interval 1438.4 to 1446.61,442(1438.4–1446.6)
  96. 96

    Arena rating

    Grok 4.3, Arena rating: 1,442 (scale 900.0 to 1526.2), 95% confidence interval 1438.0 to 1445.31,442(1438.0–1445.3)
  97. 97
    AlibabaDossier

    Arena rating

    Qwen3.5 397B A17B, Arena rating: 1,441 (scale 900.0 to 1526.2), 95% confidence interval 1438.3 to 1444.41,441(1438.3–1444.4)
  98. 98

    Arena rating

    GLM-4.7, Arena rating: 1,441 (scale 900.0 to 1526.2), 95% confidence interval 1435.3 to 1447.41,441(1435.3–1447.4)
  99. 99
    OpenAIDossier

    max

    Arena rating

    GPT-6 Luna, Arena rating: 1,441 (scale 900.0 to 1526.2), 95% confidence interval 1436.8 to 1445.01,441(1436.8–1445.0)
  100. 100
    DeepSeekDossier

    high

    Arena rating

    DeepSeek V4 Flash Preview, Arena rating: 1,439 (scale 900.0 to 1526.2), 95% confidence interval 1434.8 to 1443.11,439(1434.8–1443.1)
  101. 101
    OpenAIDossier

    Arena rating

    GPT-5.1, Arena rating: 1,438 (scale 900.0 to 1526.2), 95% confidence interval 1434.9 to 1442.11,438(1434.9–1442.1)
  102. 102
    GoogleDossier

    Arena rating

    Gemma 4 26B A4B, Arena rating: 1,437 (scale 900.0 to 1526.2), 95% confidence interval 1430.0 to 1444.91,437(1430.0–1444.9)
  103. 103
    OpenAIDossier

    high

    Arena rating

    GPT-5.2, Arena rating: 1,437 (scale 900.0 to 1526.2), 95% confidence interval 1433.7 to 1440.81,437(1433.7–1440.8)
  104. 104
    MeituanDossier

    Arena rating

    LongCat Flash Chat Exp (2602), Arena rating: 1,436 (scale 900.0 to 1526.2), 95% confidence interval 1431.4 to 1440.51,436(1431.4–1440.5)
  105. 105
    DeepSeekDossier

    Arena rating

    DeepSeek V4-Flash, Arena rating: 1,436 (scale 900.0 to 1526.2), 95% confidence interval 1431.7 to 1439.81,436(1431.7–1439.8)
  106. 106
    MiniMaxDossier

    Arena rating

    MiniMax M3, Arena rating: 1,436 (scale 900.0 to 1526.2), 95% confidence interval 1432.1 to 1439.41,436(1432.1–1439.4)
  107. 107
    OpenAIDossier

    Arena rating

    GPT-5.2, Arena rating: 1,435 (scale 900.0 to 1526.2), 95% confidence interval 1432.2 to 1438.51,435(1432.2–1438.5)
  108. 108
    OpenAIDossier

    high

    Arena rating

    GPT-5, Arena rating: 1,435 (scale 900.0 to 1526.2), 95% confidence interval 1430.4 to 1439.41,435(1430.4–1439.4)
  109. 109
    ReflectionDossier

    Arena rating

    Beam 1.0, Arena rating: 1,434 (scale 900.0 to 1526.2), 95% confidence interval 1424.8 to 1444.11,434(1424.8–1444.1)
  110. 110
    AlibabaDossier

    Arena rating

    Qwen3 Max Preview, Arena rating: 1,434 (scale 900.0 to 1526.2), 95% confidence interval 1429.9 to 1438.91,434(1429.9–1438.9)
  111. 111
    AlibabaDossier

    Arena rating

    Qwen3.8-27B, Arena rating: 1,434 (scale 900.0 to 1526.2), 95% confidence interval 1429.2 to 1438.41,434(1429.2–1438.4)
  112. 112
    XiaomiDossier

    Arena rating

    MiMo-V2.5, Arena rating: 1,434 (scale 900.0 to 1526.2), 95% confidence interval 1429.3 to 1437.81,434(1429.3–1437.8)
  113. 113
    GoogleDossier

    Arena rating

    Gemini 3.1 Flash-Lite Preview, Arena rating: 1,433 (scale 900.0 to 1526.2), 95% confidence interval 1429.4 to 1436.71,433(1429.4–1436.7)
  114. 114
    OpenAIDossier

    Arena rating

    o3, Arena rating: 1,432 (scale 900.0 to 1526.2), 95% confidence interval 1427.9 to 1435.21,432(1427.9–1435.2)
  115. 115
    Moonshot AIDossier

    instant

    Arena rating

    Kimi K2.5, Arena rating: 1,431 (scale 900.0 to 1526.2), 95% confidence interval 1424.3 to 1437.31,431(1424.3–1437.3)
  116. 116

    thinking

    Arena rating

    Grok 4.1 Fast, Arena rating: 1,431 (scale 900.0 to 1526.2), 95% confidence interval 1427.4 to 1433.81,431(1427.4–1433.8)
  117. 117
    XiaomiDossier

    Arena rating

    MiMo-V2-Omni, Arena rating: 1,430 (scale 900.0 to 1526.2), 95% confidence interval 1424.8 to 1435.91,430(1424.8–1435.9)
  118. 118
    Moonshot AIDossier

    Arena rating

    Kimi K2 Thinking Turbo, Arena rating: 1,430 (scale 900.0 to 1526.2), 95% confidence interval 1426.8 to 1433.01,430(1426.8–1433.0)
  119. 119
    OpenAIDossier

    Arena rating

    GPT-5 Chat, Arena rating: 1,427 (scale 900.0 to 1526.2), 95% confidence interval 1422.4 to 1431.01,427(1422.4–1431.0)
  120. 120
    AmazonDossier

    Arena rating

    Nova Experimental Chat (26-02-10), Arena rating: 1,427 (scale 900.0 to 1526.2), 95% confidence interval 1416.9 to 1436.31,427(1416.9–1436.3)
  121. 121
    MistralDossier

    Arena rating

    Mistral Medium 3.5, Arena rating: 1,426 (scale 900.0 to 1526.2), 95% confidence interval 1419.9 to 1432.81,426(1419.9–1432.8)
  122. 122
    AnthropicDossier

    thinking · 16,000 budget

    Arena rating

    Opus 4, Arena rating: 1,426 (scale 900.0 to 1526.2), 95% confidence interval 1421.9 to 1430.71,426(1421.9–1430.7)
  123. 123
    NVIDIADossier

    Arena rating

    Nemotron 3 Ultra 550B A55B (NVFP4), Arena rating: 1,426 (scale 900.0 to 1526.2), 95% confidence interval 1419.4 to 1433.21,426(1419.4–1433.2)
  124. 124
    DeepSeekDossier

    thinking

    Arena rating

    DeepSeek V3.2 Exp, Arena rating: 1,425 (scale 900.0 to 1526.2), 95% confidence interval 1418.5 to 1431.61,425(1418.5–1431.6)
  125. 125
    DeepSeekDossier

    Arena rating

    DeepSeek V3.2, Arena rating: 1,425 (scale 900.0 to 1526.2), 95% confidence interval 1421.2 to 1428.21,425(1421.2–1428.2)
  126. 126

    Arena rating

    Muse Glimmer, Arena rating: 1,425 (scale 900.0 to 1526.2), 95% confidence interval 1414.8 to 1434.41,425(1414.8–1434.4)
  127. 127

    Arena rating

    GLM-4.6, Arena rating: 1,424 (scale 900.0 to 1526.2), 95% confidence interval 1420.6 to 1428.41,424(1420.6–1428.4)
  128. 128
    AlibabaDossier

    Arena rating

    Qwen3 Max, Arena rating: 1,423 (scale 900.0 to 1526.2), 95% confidence interval 1416.8 to 1429.81,423(1416.8–1429.8)
  129. 129
    DeepSeekDossier

    thinking

    Arena rating

    DeepSeek V3.2, Arena rating: 1,423 (scale 900.0 to 1526.2), 95% confidence interval 1419.1 to 1426.21,423(1419.1–1426.2)
  130. 130
    AlibabaDossier

    Arena rating

    Qwen3 235B A22B 2507 Instruct, Arena rating: 1,422 (scale 900.0 to 1526.2), 95% confidence interval 1419.6 to 1424.71,422(1419.6–1424.7)
  131. 131
    DeepSeekDossier

    Arena rating

    DeepSeek V3.2 Exp, Arena rating: 1,422 (scale 900.0 to 1526.2), 95% confidence interval 1415.5 to 1428.31,422(1415.5–1428.3)
  132. 132
    DeepSeekDossier

    Arena rating

    DeepSeek R1 0528, Arena rating: 1,422 (scale 900.0 to 1526.2), 95% confidence interval 1415.8 to 1427.21,422(1415.8–1427.2)
  133. 133
    AnthropicDossier

    high

    Arena rating

    Haiku 5.5, Arena rating: 1,421 (scale 900.0 to 1526.2), 95% confidence interval 1415.0 to 1427.31,421(1415.0–1427.3)
  134. 134
    MistralDossier

    Arena rating

    Mistral Large 4, Arena rating: 1,420 (scale 900.0 to 1526.2), 95% confidence interval 1412.0 to 1428.91,420(1412.0–1428.9)
  135. 135

    non-thinking

    Arena rating

    Grok 4 Fast, Arena rating: 1,418 (scale 900.0 to 1526.2), 95% confidence interval 1410.7 to 1426.31,418(1410.7–1426.3)
  136. 136
    Moonshot AIDossier

    Arena rating

    Kimi K2 0905, Arena rating: 1,418 (scale 900.0 to 1526.2), 95% confidence interval 1411.9 to 1424.91,418(1411.9–1424.9)
  137. 137
    Moonshot AIDossier

    Arena rating

    Kimi K2, Arena rating: 1,418 (scale 900.0 to 1526.2), 95% confidence interval 1413.2 to 1423.01,418(1413.2–1423.0)
  138. 138
    BaiduDossier

    Arena rating

    ERNIE 5.0 Preview (1022), Arena rating: 1,418 (scale 900.0 to 1526.2), 95% confidence interval 1409.3 to 1426.81,418(1409.3–1426.8)
  139. 139
    DeepSeekDossier

    Arena rating

    DeepSeek V3.1, Arena rating: 1,417 (scale 900.0 to 1526.2), 95% confidence interval 1411.2 to 1423.41,417(1411.2–1423.4)
  140. 140
    AlibabaDossier

    Arena rating

    Qwen3.5 122B A10B, Arena rating: 1,416 (scale 900.0 to 1526.2), 95% confidence interval 1412.1 to 1420.71,416(1412.1–1420.7)
  141. 141
    DeepSeekDossier

    thinking

    Arena rating

    DeepSeek V3.1 Terminus, Arena rating: 1,416 (scale 900.0 to 1526.2), 95% confidence interval 1405.8 to 1426.11,416(1405.8–1426.1)
  142. 142
    DeepSeekDossier

    thinking

    Arena rating

    DeepSeek V3.1, Arena rating: 1,416 (scale 900.0 to 1526.2), 95% confidence interval 1409.2 to 1422.51,416(1409.2–1422.5)
  143. 143
    MiniMaxDossier

    Arena rating

    MiniMax M2.7, Arena rating: 1,415 (scale 900.0 to 1526.2), 95% confidence interval 1411.4 to 1418.11,415(1411.4–1418.1)
  144. 144
    DeepSeekDossier

    Arena rating

    DeepSeek V3.1 Terminus, Arena rating: 1,415 (scale 900.0 to 1526.2), 95% confidence interval 1405.0 to 1424.51,415(1405.0–1424.5)
  145. 145
    MistralDossier

    Arena rating

    Mistral Large 3, Arena rating: 1,415 (scale 900.0 to 1526.2), 95% confidence interval 1411.9 to 1417.51,415(1411.9–1417.5)
  146. 146
    OpenAIDossier

    Arena rating

    GPT-4.1, Arena rating: 1,415 (scale 900.0 to 1526.2), 95% confidence interval 1410.9 to 1418.41,415(1410.9–1418.4)
  147. 147
    AnthropicDossier

    Arena rating

    Opus 4, Arena rating: 1,414 (scale 900.0 to 1526.2), 95% confidence interval 1410.1 to 1418.81,414(1410.1–1418.8)
  148. 148
    NVIDIADossier

    Arena rating

    Nemotron 3.5 Super VL 120B A12B (NVFP4), Arena rating: 1,414 (scale 900.0 to 1526.2), 95% confidence interval 1404.8 to 1424.01,414(1404.8–1424.0)
  149. 149
    AnthropicDossier

    Arena rating

    Haiku 4.5, Arena rating: 1,414 (scale 900.0 to 1526.2), 95% confidence interval 1411.6 to 1416.41,414(1411.6–1416.4)
  150. 150
    AlibabaDossier

    Arena rating

    Qwen3 VL 235B A22B Instruct, Arena rating: 1,413 (scale 900.0 to 1526.2), 95% confidence interval 1406.6 to 1419.71,413(1406.6–1419.7)
  151. 151
    AmazonDossier

    Arena rating

    Nova Experimental Chat (26-01-10), Arena rating: 1,412 (scale 900.0 to 1526.2), 95% confidence interval 1402.1 to 1421.51,412(1402.1–1421.5)
  152. 152
    Thinking MachinesDossier

    Arena rating

    Inkling 2 Small, Arena rating: 1,412 (scale 900.0 to 1526.2), 95% confidence interval 1400.0 to 1423.61,412(1400.0–1423.6)
  153. 153

    Arena rating

    Grok 3 Preview (02-24), Arena rating: 1,411 (scale 900.0 to 1526.2), 95% confidence interval 1407.0 to 1415.71,411(1407.0–1415.7)
  154. 154

    Arena rating

    Grok 4 (0709), Arena rating: 1,411 (scale 900.0 to 1526.2), 95% confidence interval 1407.1 to 1415.01,411(1407.1–1415.0)
  155. 155

    Arena rating

    GLM-4.5, Arena rating: 1,411 (scale 900.0 to 1526.2), 95% confidence interval 1406.0 to 1415.91,411(1406.0–1415.9)
  156. 156
    GoogleDossier

    Arena rating

    Gemini 2.5 Flash, Arena rating: 1,409 (scale 900.0 to 1526.2), 95% confidence interval 1406.9 to 1411.81,409(1406.9–1411.8)
  157. 157
    TencentDossier

    Arena rating

    Hy3 Preview, Arena rating: 1,409 (scale 900.0 to 1526.2), 95% confidence interval 1401.4 to 1416.31,409(1401.4–1416.3)
  158. 158
    AlibabaDossier

    Arena rating

    Qwen3.5 27B, Arena rating: 1,408 (scale 900.0 to 1526.2), 95% confidence interval 1404.1 to 1412.81,408(1404.1–1412.8)
  159. 159
    MistralDossier

    Arena rating

    Mistral Medium 3.1, Arena rating: 1,408 (scale 900.0 to 1526.2), 95% confidence interval 1405.5 to 1410.71,408(1405.5–1410.7)
  160. 160
    Thinking MachinesDossier

    Arena rating

    Inkling Small, Arena rating: 1,405 (scale 900.0 to 1526.2), 95% confidence interval 1400.9 to 1409.91,405(1400.9–1409.9)
  161. 161

    thinking

    Arena rating

    Grok 4 Fast, Arena rating: 1,405 (scale 900.0 to 1526.2), 95% confidence interval 1400.1 to 1410.31,405(1400.1–1410.3)
  162. 162
    GoogleDossier

    Arena rating

    Gemini 2.5 Flash Preview (09-2025), Arena rating: 1,403 (scale 900.0 to 1526.2), 95% confidence interval 1399.1 to 1407.21,403(1399.1–1407.2)
  163. 163
    OpenAIDossier

    Arena rating

    o1, Arena rating: 1,402 (scale 900.0 to 1526.2), 95% confidence interval 1397.9 to 1406.81,402(1397.9–1406.8)
  164. 164
    AlibabaDossier

    non-thinking

    Arena rating

    Qwen3 235B A22B, Arena rating: 1,402 (scale 900.0 to 1526.2), 95% confidence interval 1397.7 to 1406.71,402(1397.7–1406.7)
  165. 165
    AnthropicDossier

    thinking · 32,000 budget

    Arena rating

    Claude Sonnet 4, Arena rating: 1,402 (scale 900.0 to 1526.2), 95% confidence interval 1397.5 to 1406.41,402(1397.5–1406.4)
  166. 166
    OpenAIDossier

    high

    Arena rating

    GPT-5.4 nano, Arena rating: 1,401 (scale 900.0 to 1526.2), 95% confidence interval 1397.2 to 1404.81,401(1397.2–1404.8)
  167. 167
    MeituanDossier

    Arena rating

    LongCat Flash Chat, Arena rating: 1,401 (scale 900.0 to 1526.2), 95% confidence interval 1394.2 to 1407.01,401(1394.2–1407.0)
  168. 168
    AlibabaDossier

    Arena rating

    Qwen3 235B A22B Thinking (2507), Arena rating: 1,400 (scale 900.0 to 1526.2), 95% confidence interval 1393.6 to 1406.81,400(1393.6–1406.8)
  169. 169
    AlibabaDossier

    Arena rating

    Qwen3 Next 80B A3B Instruct, Arena rating: 1,399 (scale 900.0 to 1526.2), 95% confidence interval 1393.8 to 1403.41,399(1393.8–1403.4)
  170. 170
    DeepSeekDossier

    Arena rating

    DeepSeek R1 (Jan '25), Arena rating: 1,398 (scale 900.0 to 1526.2), 95% confidence interval 1393.3 to 1403.11,398(1393.3–1403.1)
  171. 171
    AlibabaDossier

    Arena rating

    Qwen3.5 Flash, Arena rating: 1,396 (scale 900.0 to 1526.2), 95% confidence interval 1392.3 to 1399.61,396(1392.3–1399.6)
  172. 172
    DeepSeekDossier

    Arena rating

    DeepSeek V3 0324, Arena rating: 1,395 (scale 900.0 to 1526.2), 95% confidence interval 1391.4 to 1399.21,395(1391.4–1399.2)
  173. 173
    AlibabaDossier

    Arena rating

    Qwen3 VL 235B A22B Thinking, Arena rating: 1,395 (scale 900.0 to 1526.2), 95% confidence interval 1387.8 to 1401.51,395(1387.8–1401.5)
  174. 174
    TencentDossier

    Arena rating

    Hunyuan Vision 1.5 Thinking, Arena rating: 1,395 (scale 900.0 to 1526.2), 95% confidence interval 1382.1 to 1407.01,395(1382.1–1407.0)
  175. 175
    AlibabaDossier

    Arena rating

    Qwen3.5 35B A3B, Arena rating: 1,394 (scale 900.0 to 1526.2), 95% confidence interval 1389.7 to 1398.21,394(1389.7–1398.2)
  176. 176
    StepFunDossier

    Arena rating

    Step 3.5 Flash, Arena rating: 1,393 (scale 900.0 to 1526.2), 95% confidence interval 1389.0 to 1396.11,393(1389.0–1396.1)
  177. 177
    AmazonDossier

    Arena rating

    Nova Experimental Chat (12-10), Arena rating: 1,393 (scale 900.0 to 1526.2), 95% confidence interval 1383.0 to 1402.01,393(1383.0–1402.0)
  178. 178
    XiaomiDossier

    non-thinking

    Arena rating

    MiMo-V2-Flash, Arena rating: 1,392 (scale 900.0 to 1526.2), 95% confidence interval 1388.2 to 1395.21,392(1388.2–1395.2)
  179. 179
    AnthropicDossier

    Arena rating

    Claude Sonnet 4, Arena rating: 1,391 (scale 900.0 to 1526.2), 95% confidence interval 1386.5 to 1395.31,391(1386.5–1395.3)
  180. 180
    OpenAIDossier

    Arena rating

    o4-mini, Arena rating: 1,391 (scale 900.0 to 1526.2), 95% confidence interval 1386.8 to 1394.81,391(1386.8–1394.8)
  181. 181
    MiniMaxDossier

    Arena rating

    MiniMax M2.5, Arena rating: 1,391 (scale 900.0 to 1526.2), 95% confidence interval 1386.9 to 1394.71,391(1386.9–1394.7)
  182. 182
    OpenAIDossier

    high

    Arena rating

    GPT-5 mini, Arena rating: 1,390 (scale 900.0 to 1526.2), 95% confidence interval 1385.2 to 1394.51,390(1385.2–1394.5)
  183. 183
    OpenAIDossier

    Arena rating

    o1-preview, Arena rating: 1,389 (scale 900.0 to 1526.2), 95% confidence interval 1383.7 to 1393.71,389(1383.7–1393.7)
  184. 184
    AnthropicDossier

    thinking · 32,000 budget

    Arena rating

    Claude 3.7 Sonnet, Arena rating: 1,388 (scale 900.0 to 1526.2), 95% confidence interval 1384.1 to 1392.61,388(1384.1–1392.6)
  185. 185
    UpstageDossier

    Arena rating

    Solar Pro 4, Arena rating: 1,387 (scale 900.0 to 1526.2), 95% confidence interval 1382.4 to 1392.51,387(1382.4–1392.5)
  186. 186
    AlibabaDossier

    Arena rating

    Qwen3 Coder 480B A35B Instruct, Arena rating: 1,387 (scale 900.0 to 1526.2), 95% confidence interval 1382.5 to 1392.41,387(1382.5–1392.4)
  187. 187
    MistralDossier

    Arena rating

    Mistral Medium 3, Arena rating: 1,387 (scale 900.0 to 1526.2), 95% confidence interval 1382.6 to 1392.11,387(1382.6–1392.1)
  188. 188
    TencentDossier

    Arena rating

    Hunyuan T1, Arena rating: 1,387 (scale 900.0 to 1526.2), 95% confidence interval 1378.6 to 1396.01,387(1378.6–1396.0)
  189. 189
    XiaomiDossier

    thinking

    Arena rating

    MiMo-V2-Flash, Arena rating: 1,386 (scale 900.0 to 1526.2), 95% confidence interval 1380.0 to 1392.41,386(1380.0–1392.4)
  190. 190
    MiniMaxDossier

    Arena rating

    MiniMax M2.1 Preview, Arena rating: 1,384 (scale 900.0 to 1526.2), 95% confidence interval 1379.2 to 1389.51,384(1379.2–1389.5)
  191. 191
    TencentDossier

    Arena rating

    Hunyuan TurboS, Arena rating: 1,383 (scale 900.0 to 1526.2), 95% confidence interval 1376.6 to 1389.61,383(1376.6–1389.6)
  192. 192
    OpenAIDossier

    Arena rating

    GPT-4.1 mini, Arena rating: 1,383 (scale 900.0 to 1526.2), 95% confidence interval 1378.6 to 1387.31,383(1378.6–1387.3)
  193. 193
    AlibabaDossier

    Arena rating

    Qwen3 30B A3B 2507 Instruct, Arena rating: 1,382 (scale 900.0 to 1526.2), 95% confidence interval 1377.4 to 1387.21,382(1377.4–1387.2)
  194. 194
    GoogleDossier

    non-thinking

    Arena rating

    Gemini 2.5 Flash-Lite Preview (Sep '25), Arena rating: 1,379 (scale 900.0 to 1526.2), 95% confidence interval 1376.0 to 1382.81,379(1376.0–1382.8)
  195. 195
    Arcee AIDossier

    Arena rating

    Trinity Large Preview, Arena rating: 1,379 (scale 900.0 to 1526.2), 95% confidence interval 1374.4 to 1383.01,379(1374.4–1383.0)
  196. 196

    Arena rating

    GLM-4.6V, Arena rating: 1,378 (scale 900.0 to 1526.2), 95% confidence interval 1367.1 to 1389.61,378(1367.1–1389.6)
  197. 197
    AlibabaDossier

    Arena rating

    Qwen3 235B A22B, Arena rating: 1,375 (scale 900.0 to 1526.2), 95% confidence interval 1370.2 to 1379.61,375(1370.2–1379.6)
  198. 198
    GoogleDossier

    thinking

    Arena rating

    Gemini 2.5 Flash-Lite Preview (06-17), Arena rating: 1,375 (scale 900.0 to 1526.2), 95% confidence interval 1370.4 to 1379.41,375(1370.4–1379.4)
  199. 199
    AnthropicDossier

    Arena rating

    Claude 3.5 Sonnet (Oct '24), Arena rating: 1,374 (scale 900.0 to 1526.2), 95% confidence interval 1371.1 to 1377.51,374(1371.1–1377.5)
  200. 200
    AlibabaDossier

    Arena rating

    Qwen2.5 Max, Arena rating: 1,374 (scale 900.0 to 1526.2), 95% confidence interval 1369.7 to 1377.91,374(1369.7–1377.9)
  201. 201

    Arena rating

    GLM-4.5-Air, Arena rating: 1,373 (scale 900.0 to 1526.2), 95% confidence interval 1368.7 to 1377.31,373(1368.7–1377.3)
  202. 202
    AnthropicDossier

    Arena rating

    Claude 3.7 Sonnet, Arena rating: 1,373 (scale 900.0 to 1526.2), 95% confidence interval 1368.5 to 1376.51,373(1368.5–1376.5)
  203. 203
    AlibabaDossier

    Arena rating

    Qwen3 Next 80B A3B Thinking, Arena rating: 1,369 (scale 900.0 to 1526.2), 95% confidence interval 1363.6 to 1375.31,369(1363.6–1375.3)
  204. 204
    Arcee AIDossier

    Arena rating

    Trinity Large Thinking, Arena rating: 1,367 (scale 900.0 to 1526.2), 95% confidence interval 1362.6 to 1371.81,367(1362.6–1371.8)
  205. 205
    GoogleDossier

    Arena rating

    Gemma 3 27B, Arena rating: 1,365 (scale 900.0 to 1526.2), 95% confidence interval 1361.7 to 1369.11,365(1361.7–1369.1)
  206. 206

    Arena rating

    GLM-4.7-Flash, Arena rating: 1,365 (scale 900.0 to 1526.2), 95% confidence interval 1358.8 to 1370.31,365(1358.8–1370.3)
  207. 207

    high

    Arena rating

    Grok 3 mini, Arena rating: 1,364 (scale 900.0 to 1526.2), 95% confidence interval 1358.3 to 1369.31,364(1358.3–1369.3)
  208. 208
    MiniMaxDossier

    Arena rating

    MiniMax M1, Arena rating: 1,364 (scale 900.0 to 1526.2), 95% confidence interval 1359.5 to 1368.11,364(1359.5–1368.1)
  209. 209
    OpenAIDossier

    high

    Arena rating

    o3-mini, Arena rating: 1,364 (scale 900.0 to 1526.2), 95% confidence interval 1358.4 to 1368.91,364(1358.4–1368.9)
  210. 210
    AmazonDossier

    Arena rating

    Nova Experimental Chat (11-10), Arena rating: 1,363 (scale 900.0 to 1526.2), 95% confidence interval 1359.1 to 1367.71,363(1359.1–1367.7)
  211. 211
    GoogleDossier

    Arena rating

    Gemini 2.0 Flash (001), Arena rating: 1,360 (scale 900.0 to 1526.2), 95% confidence interval 1356.4 to 1364.01,360(1356.4–1364.0)
  212. 212
    NVIDIADossier

    Arena rating

    Nemotron 3 Super 120B A12B, Arena rating: 1,360 (scale 900.0 to 1526.2), 95% confidence interval 1352.5 to 1366.81,360(1352.5–1366.8)
  213. 213
    DeepSeekDossier

    Arena rating

    DeepSeek V3, Arena rating: 1,358 (scale 900.0 to 1526.2), 95% confidence interval 1353.7 to 1363.21,358(1353.7–1363.2)
  214. 214

    Arena rating

    Grok 3 Mini Beta, Arena rating: 1,358 (scale 900.0 to 1526.2), 95% confidence interval 1352.8 to 1363.11,358(1352.8–1363.1)
  215. 215
    MistralDossier

    Arena rating

    Mistral Small 3.2, Arena rating: 1,357 (scale 900.0 to 1526.2), 95% confidence interval 1351.3 to 1361.81,357(1351.3–1361.8)
  216. 216
    Prime IntellectDossier

    Arena rating

    INTELLECT-3, Arena rating: 1,356 (scale 900.0 to 1526.2), 95% confidence interval 1347.6 to 1364.11,356(1347.6–1364.1)
  217. 217
    CohereDossier

    Arena rating

    Command A, Arena rating: 1,354 (scale 900.0 to 1526.2), 95% confidence interval 1350.2 to 1357.21,354(1350.2–1357.2)
  218. 218
    GoogleDossier

    Arena rating

    Gemini 2.0 Flash-Lite Preview (02-05), Arena rating: 1,354 (scale 900.0 to 1526.2), 95% confidence interval 1349.3 to 1358.01,354(1349.3–1358.0)
  219. 219

    Arena rating

    GLM-4.5V, Arena rating: 1,352 (scale 900.0 to 1526.2), 95% confidence interval 1343.8 to 1360.71,352(1343.8–1360.7)
  220. 220
    OpenAIDossier

    Arena rating

    gpt-oss-120b, Arena rating: 1,352 (scale 900.0 to 1526.2), 95% confidence interval 1347.4 to 1356.21,352(1347.4–1356.2)
  221. 221
    NVIDIADossier

    Arena rating

    Nemotron 3.5 Lightning 30B A3B (NVFP4), Arena rating: 1,351 (scale 900.0 to 1526.2), 95% confidence interval 1346.2 to 1356.81,351(1346.2–1356.8)
  222. 222
    GoogleDossier

    Arena rating

    Gemini 1.5 Pro (002), Arena rating: 1,351 (scale 900.0 to 1526.2), 95% confidence interval 1348.0 to 1354.81,351(1348.0–1354.8)
  223. 223
    TencentDossier

    Arena rating

    Hunyuan TurboS, Arena rating: 1,349 (scale 900.0 to 1526.2), 95% confidence interval 1337.2 to 1360.61,349(1337.2–1360.6)
  224. 224
    StepFunDossier

    Arena rating

    Step-3, Arena rating: 1,349 (scale 900.0 to 1526.2), 95% confidence interval 1341.2 to 1356.31,349(1341.2–1356.3)
  225. 225
    OpenAIDossier

    Arena rating

    o3-mini, Arena rating: 1,348 (scale 900.0 to 1526.2), 95% confidence interval 1344.7 to 1351.81,348(1344.7–1351.8)
  226. 226
    AmazonDossier

    Arena rating

    Nova Experimental Chat (10-20), Arena rating: 1,348 (scale 900.0 to 1526.2), 95% confidence interval 1341.8 to 1354.11,348(1341.8–1354.1)
  227. 227
    NVIDIADossier

    Arena rating

    Llama 3.1 Nemotron Ultra 253B v1, Arena rating: 1,348 (scale 900.0 to 1526.2), 95% confidence interval 1336.0 to 1359.31,348(1336.0–1359.3)
  228. 228
    AlibabaDossier

    Arena rating

    Qwen3 32B, Arena rating: 1,347 (scale 900.0 to 1526.2), 95% confidence interval 1337.4 to 1356.31,347(1337.4–1356.3)
  229. 229
    OpenAIDossier

    Arena rating

    GPT-4o (May '24), Arena rating: 1,346 (scale 900.0 to 1526.2), 95% confidence interval 1342.7 to 1349.71,346(1342.7–1349.7)
  230. 230
    AmazonDossier

    Arena rating

    Nova Experimental Chat (10-09), Arena rating: 1,346 (scale 900.0 to 1526.2), 95% confidence interval 1335.2 to 1357.11,346(1335.2–1357.1)
  231. 231
    AlibabaDossier

    Arena rating

    Qwen Plus (0125), Arena rating: 1,346 (scale 900.0 to 1526.2), 95% confidence interval 1337.6 to 1354.21,346(1337.6–1354.2)
  232. 232
    MiniMaxDossier

    Arena rating

    MiniMax M2, Arena rating: 1,344 (scale 900.0 to 1526.2), 95% confidence interval 1336.0 to 1351.51,344(1336.0–1351.5)
  233. 233
    InceptionDossier

    Arena rating

    Mercury 2, Arena rating: 1,343 (scale 900.0 to 1526.2), 95% confidence interval 1332.8 to 1354.01,343(1332.8–1354.0)
  234. 234
    AnthropicDossier

    Arena rating

    Claude 3.5 Sonnet (June '24), Arena rating: 1,343 (scale 900.0 to 1526.2), 95% confidence interval 1339.9 to 1346.81,343(1339.9–1346.8)
  235. 235
    NVIDIADossier

    Arena rating

    Llama 3.3 Nemotron Super 49B v1.5, Arena rating: 1,343 (scale 900.0 to 1526.2), 95% confidence interval 1333.3 to 1353.31,343(1333.3–1353.3)
  236. 236
    InclusionAIDossier

    Arena rating

    Ling Flash 2.0, Arena rating: 1,343 (scale 900.0 to 1526.2), 95% confidence interval 1335.8 to 1350.41,343(1335.8–1350.4)
  237. 237

    Arena rating

    GLM-4-Plus (0111), Arena rating: 1,343 (scale 900.0 to 1526.2), 95% confidence interval 1334.4 to 1351.21,343(1334.4–1351.2)
  238. 238
    GoogleDossier

    Arena rating

    Gemma 3 12B, Arena rating: 1,342 (scale 900.0 to 1526.2), 95% confidence interval 1332.3 to 1351.31,342(1332.3–1351.3)
  239. 239
    TencentDossier

    Arena rating

    Hunyuan Turbo (0110), Arena rating: 1,341 (scale 900.0 to 1526.2), 95% confidence interval 1329.2 to 1352.31,341(1329.2–1352.3)
  240. 240

    Arena rating

    Granite 4.2 30B, Arena rating: 1,340 (scale 900.0 to 1526.2), 95% confidence interval 1329.5 to 1350.71,340(1329.5–1350.7)
  241. 241
    OpenAIDossier

    high

    Arena rating

    GPT-5 nano, Arena rating: 1,338 (scale 900.0 to 1526.2), 95% confidence interval 1330.6 to 1344.51,338(1330.6–1344.5)
  242. 242
    OpenAIDossier

    Arena rating

    o1-mini, Arena rating: 1,337 (scale 900.0 to 1526.2), 95% confidence interval 1333.2 to 1340.51,337(1333.2–1340.5)
  243. 243
    GoogleDossier

    Arena rating

    Gemini Advanced (0514), Arena rating: 1,336 (scale 900.0 to 1526.2), 95% confidence interval 1331.0 to 1341.31,336(1331.0–1341.3)
  244. 244

    Arena rating

    Grok 2, Arena rating: 1,336 (scale 900.0 to 1526.2), 95% confidence interval 1332.0 to 1339.41,336(1332.0–1339.4)
  245. 245
    OpenAIDossier

    Arena rating

    GPT-4o (Aug '24), Arena rating: 1,336 (scale 900.0 to 1526.2), 95% confidence interval 1331.5 to 1339.91,336(1331.5–1339.9)
  246. 246

    Arena rating

    Llama 3.1 405B Instruct (BF16), Arena rating: 1,336 (scale 900.0 to 1526.2), 95% confidence interval 1331.8 to 1339.31,336(1331.8–1339.3)
  247. 247
    AlibabaDossier

    Arena rating

    QwQ 32B, Arena rating: 1,336 (scale 900.0 to 1526.2), 95% confidence interval 1331.0 to 1340.01,336(1331.0–1340.0)
  248. 248
    AmazonDossier

    Arena rating

    Nova 2.0 Lite, Arena rating: 1,335 (scale 900.0 to 1526.2), 95% confidence interval 1328.7 to 1340.91,335(1328.7–1340.9)
  249. 249

    Arena rating

    Llama 3.1 405B Instruct (FP8), Arena rating: 1,334 (scale 900.0 to 1526.2), 95% confidence interval 1330.1 to 1337.31,334(1330.1–1337.3)
  250. 250
    StepFunDossier

    Arena rating

    Step-2 16K Exp (202412), Arena rating: 1,334 (scale 900.0 to 1526.2), 95% confidence interval 1325.0 to 1342.11,334(1325.0–1342.1)
  251. 251
    Allen Institute for AIDossier

    Arena rating

    Olmo 3.1 32B Instruct, Arena rating: 1,329 (scale 900.0 to 1526.2), 95% confidence interval 1323.2 to 1335.51,329(1323.2–1335.5)
  252. 252
    01.AIDossier

    Arena rating

    Yi Lightning, Arena rating: 1,328 (scale 900.0 to 1526.2), 95% confidence interval 1323.4 to 1333.31,328(1323.4–1333.3)
  253. 253
    NVIDIADossier

    Arena rating

    Llama 3.3 Nemotron Super 49B v1, Arena rating: 1,328 (scale 900.0 to 1526.2), 95% confidence interval 1315.7 to 1339.91,328(1315.7–1339.9)
  254. 254

    Arena rating

    Llama 4 Maverick, Arena rating: 1,327 (scale 900.0 to 1526.2), 95% confidence interval 1322.8 to 1331.31,327(1322.8–1331.3)
  255. 255
    AlibabaDossier

    Arena rating

    Qwen3 30B A3B, Arena rating: 1,327 (scale 900.0 to 1526.2), 95% confidence interval 1321.8 to 1331.31,327(1321.8–1331.3)
  256. 256
    TencentDossier

    Arena rating

    Hunyuan Large, Arena rating: 1,326 (scale 900.0 to 1526.2), 95% confidence interval 1316.7 to 1336.11,326(1316.7–1336.1)
  257. 257
    AnthropicDossier

    Arena rating

    Claude 3.5 Haiku, Arena rating: 1,325 (scale 900.0 to 1526.2), 95% confidence interval 1321.5 to 1328.01,325(1321.5–1328.0)
  258. 258
    OpenAIDossier

    Arena rating

    GPT-4 Turbo, Arena rating: 1,325 (scale 900.0 to 1526.2), 95% confidence interval 1320.7 to 1328.51,325(1320.7–1328.5)
  259. 259
    GoogleDossier

    Arena rating

    Gemini 1.5 Pro (001), Arena rating: 1,324 (scale 900.0 to 1526.2), 95% confidence interval 1320.2 to 1328.21,324(1320.2–1328.2)
  260. 260
    DeepSeekDossier

    Arena rating

    DeepSeek V2.5 (1210), Arena rating: 1,323 (scale 900.0 to 1526.2), 95% confidence interval 1315.2 to 1331.71,323(1315.2–1331.7)
  261. 261
    OpenAIDossier

    Arena rating

    GPT-4.1 nano, Arena rating: 1,322 (scale 900.0 to 1526.2), 95% confidence interval 1314.6 to 1330.11,322(1314.6–1330.1)
  262. 262
    AnthropicDossier

    Arena rating

    Claude 3 Opus, Arena rating: 1,322 (scale 900.0 to 1526.2), 95% confidence interval 1319.2 to 1325.41,322(1319.2–1325.4)
  263. 263
    InclusionAIDossier

    Arena rating

    Ring-flash-2.0, Arena rating: 1,322 (scale 900.0 to 1526.2), 95% confidence interval 1314.9 to 1329.51,322(1314.9–1329.5)
  264. 264

    Arena rating

    Llama 4 Scout, Arena rating: 1,322 (scale 900.0 to 1526.2), 95% confidence interval 1316.9 to 1326.31,322(1316.9–1326.3)
  265. 265
    Allen Institute for AIDossier

    Arena rating

    Molmo 2 8B, Arena rating: 1,321 (scale 900.0 to 1526.2), 95% confidence interval 1299.9 to 1342.21,321(1299.9–1342.2)
  266. 266
    StepFunDossier

    Arena rating

    Step-1o Turbo (202506), Arena rating: 1,320 (scale 900.0 to 1526.2), 95% confidence interval 1312.8 to 1326.51,320(1312.8–1326.5)
  267. 267

    Arena rating

    GLM-4-Plus, Arena rating: 1,319 (scale 900.0 to 1526.2), 95% confidence interval 1314.5 to 1324.31,319(1314.5–1324.3)
  268. 268
    AlibabaDossier

    Arena rating

    Qwen Max (0919), Arena rating: 1,318 (scale 900.0 to 1526.2), 95% confidence interval 1312.5 to 1323.91,318(1312.5–1323.9)
  269. 269

    Arena rating

    Llama 3.3 Instruct 70B, Arena rating: 1,318 (scale 900.0 to 1526.2), 95% confidence interval 1314.6 to 1321.71,318(1314.6–1321.7)
  270. 270
    OpenAIDossier

    Arena rating

    GPT-4o mini, Arena rating: 1,318 (scale 900.0 to 1526.2), 95% confidence interval 1314.1 to 1321.21,318(1314.1–1321.2)
  271. 271
    GoogleDossier

    Arena rating

    Gemma 3n E4B IT, Arena rating: 1,318 (scale 900.0 to 1526.2), 95% confidence interval 1312.4 to 1322.81,318(1312.4–1322.8)
  272. 272
    OpenAIDossier

    Arena rating

    gpt-oss-20b, Arena rating: 1,318 (scale 900.0 to 1526.2), 95% confidence interval 1311.1 to 1323.91,318(1311.1–1323.9)
  273. 273
    AlibabaDossier

    Arena rating

    Qwen2.5 Plus (1127), Arena rating: 1,315 (scale 900.0 to 1526.2), 95% confidence interval 1308.3 to 1321.01,315(1308.3–1321.0)
  274. 274
    MistralDossier

    Arena rating

    Mistral Large 2 (Jul '24), Arena rating: 1,315 (scale 900.0 to 1526.2), 95% confidence interval 1310.6 to 1318.41,315(1310.6–1318.4)
  275. 275
    NexusflowDossier

    Arena rating

    Athene V2 Chat, Arena rating: 1,314 (scale 900.0 to 1526.2), 95% confidence interval 1309.6 to 1318.71,314(1309.6–1318.7)
  276. 276
    NVIDIADossier

    Arena rating

    Nemotron 3 Nano 30B A3B (BF16), Arena rating: 1,314 (scale 900.0 to 1526.2), 95% confidence interval 1308.2 to 1319.21,314(1308.2–1319.2)
  277. 277
    OpenAIDossier

    Arena rating

    GPT-4 Turbo Preview (0125), Arena rating: 1,313 (scale 900.0 to 1526.2), 95% confidence interval 1309.3 to 1317.41,313(1309.3–1317.4)
  278. 278
    OpenAIDossier

    Arena rating

    GPT-4 Turbo Preview (1106), Arena rating: 1,313 (scale 900.0 to 1526.2), 95% confidence interval 1309.1 to 1316.91,313(1309.1–1316.9)
  279. 279
    TencentDossier

    Arena rating

    Hunyuan Standard, Arena rating: 1,311 (scale 900.0 to 1526.2), 95% confidence interval 1301.7 to 1321.01,311(1301.7–1321.0)
  280. 280
    GoogleDossier

    Arena rating

    Gemini 1.5 Flash (002), Arena rating: 1,309 (scale 900.0 to 1526.2), 95% confidence interval 1305.2 to 1313.61,309(1305.2–1313.6)
  281. 281

    Arena rating

    Grok 2 mini, Arena rating: 1,308 (scale 900.0 to 1526.2), 95% confidence interval 1304.7 to 1312.21,308(1304.7–1312.2)
  282. 282
    DeepSeekDossier

    Arena rating

    DeepSeek V2.5, Arena rating: 1,307 (scale 900.0 to 1526.2), 95% confidence interval 1302.5 to 1312.01,307(1302.5–1312.0)
  283. 283
    NexusflowDossier

    Arena rating

    Athene 70B (0725), Arena rating: 1,307 (scale 900.0 to 1526.2), 95% confidence interval 1301.1 to 1312.41,307(1301.1–1312.4)
  284. 284
    Allen Institute for AIDossier

    Arena rating

    Olmo 3 32B Think, Arena rating: 1,306 (scale 900.0 to 1526.2), 95% confidence interval 1297.5 to 1314.01,306(1297.5–1314.0)
  285. 285
    MistralDossier

    Arena rating

    Mistral Large (2411), Arena rating: 1,306 (scale 900.0 to 1526.2), 95% confidence interval 1301.2 to 1310.11,306(1301.2–1310.1)
  286. 286
    InceptionDossier

    Arena rating

    Mercury, Arena rating: 1,305 (scale 900.0 to 1526.2), 95% confidence interval 1291.5 to 1319.11,305(1291.5–1319.1)
  287. 287
    MistralDossier

    Arena rating

    Magistral Medium (2506), Arena rating: 1,305 (scale 900.0 to 1526.2), 95% confidence interval 1298.3 to 1311.31,305(1298.3–1311.3)
  288. 288

    Arena rating

    Granite 4.1 8B, Arena rating: 1,304 (scale 900.0 to 1526.2), 95% confidence interval 1294.3 to 1314.31,304(1294.3–1314.3)
  289. 289
    GoogleDossier

    Arena rating

    Gemma 3 4B, Arena rating: 1,303 (scale 900.0 to 1526.2), 95% confidence interval 1294.0 to 1312.61,303(1294.0–1312.6)
  290. 290
    AlibabaDossier

    Arena rating

    Qwen2.5 72B Instruct, Arena rating: 1,303 (scale 900.0 to 1526.2), 95% confidence interval 1298.8 to 1307.01,303(1298.8–1307.0)
  291. 291
    MistralDossier

    Arena rating

    Mistral Small 3.1, Arena rating: 1,303 (scale 900.0 to 1526.2), 95% confidence interval 1298.2 to 1307.31,303(1298.2–1307.3)
  292. 292
    NVIDIADossier

    Arena rating

    Llama 3.1 Nemotron 70B Instruct, Arena rating: 1,299 (scale 900.0 to 1526.2), 95% confidence interval 1291.0 to 1306.61,299(1291.0–1306.6)
  293. 293
    TencentDossier

    Arena rating

    Hunyuan Large Vision, Arena rating: 1,294 (scale 900.0 to 1526.2), 95% confidence interval 1285.0 to 1303.31,294(1285.0–1303.3)
  294. 294

    Arena rating

    Llama 3.1 Instruct 70B, Arena rating: 1,293 (scale 900.0 to 1526.2), 95% confidence interval 1289.7 to 1297.21,293(1289.7–1297.2)
  295. 295
    AmazonDossier

    Arena rating

    Nova Pro, Arena rating: 1,290 (scale 900.0 to 1526.2), 95% confidence interval 1285.6 to 1294.71,290(1285.6–1294.7)
  296. 296
    GoogleDossier

    Arena rating

    Gemma 2 27B IT, Arena rating: 1,290 (scale 900.0 to 1526.2), 95% confidence interval 1286.2 to 1293.01,290(1286.2–1293.0)
  297. 297
    AI21 LabsDossier

    Arena rating

    Jamba 1.5 Large, Arena rating: 1,290 (scale 900.0 to 1526.2), 95% confidence interval 1282.3 to 1296.91,290(1282.3–1296.9)
  298. 298

    Arena rating

    Granite 4.2 8B, Arena rating: 1,289 (scale 900.0 to 1526.2), 95% confidence interval 1278.0 to 1300.61,289(1278.0–1300.6)
  299. 299

    Arena rating

    Granite 4.2 3B, Arena rating: 1,289 (scale 900.0 to 1526.2), 95% confidence interval 1277.2 to 1300.71,289(1277.2–1300.7)
  300. 300
    Reka AIDossier

    Arena rating

    Reka Core, Arena rating: 1,288 (scale 900.0 to 1526.2), 95% confidence interval 1280.9 to 1295.31,288(1280.9–1295.3)
  301. 301
    OpenAIDossier

    Arena rating

    GPT-4 (0314), Arena rating: 1,288 (scale 900.0 to 1526.2), 95% confidence interval 1283.3 to 1292.81,288(1283.3–1292.8)
  302. 302

    Arena rating

    Granite 4.0 H Small, Arena rating: 1,288 (scale 900.0 to 1526.2), 95% confidence interval 1279.1 to 1295.91,288(1279.1–1295.9)
  303. 303
    NVIDIADossier

    Arena rating

    Llama 3.1 Nemotron 51B Instruct, Arena rating: 1,287 (scale 900.0 to 1526.2), 95% confidence interval 1276.8 to 1296.71,287(1276.8–1296.7)
  304. 304
    GoogleDossier

    Arena rating

    Gemini 1.5 Flash (001), Arena rating: 1,287 (scale 900.0 to 1526.2), 95% confidence interval 1282.2 to 1291.21,287(1282.2–1291.2)
  305. 305
    Allen Institute for AIDossier

    Arena rating

    Llama 3.1 Tulu 3 70B, Arena rating: 1,286 (scale 900.0 to 1526.2), 95% confidence interval 1275.7 to 1296.61,286(1275.7–1296.6)
  306. 306
    Allen Institute for AIDossier

    Arena rating

    Olmo 3.1 32B Think, Arena rating: 1,286 (scale 900.0 to 1526.2), 95% confidence interval 1278.8 to 1293.41,286(1278.8–1293.4)
  307. 307
    AnthropicDossier

    Arena rating

    Claude 3 Sonnet, Arena rating: 1,281 (scale 900.0 to 1526.2), 95% confidence interval 1277.5 to 1285.51,281(1277.5–1285.5)
  308. 308
    Princeton NLPDossier

    Arena rating

    Gemma 2 9B IT SimPO, Arena rating: 1,281 (scale 900.0 to 1526.2), 95% confidence interval 1273.7 to 1287.41,281(1273.7–1287.4)
  309. 309
    NVIDIADossier

    Arena rating

    Nemotron-4 340B Instruct, Arena rating: 1,277 (scale 900.0 to 1526.2), 95% confidence interval 1272.0 to 1282.61,277(1272.0–1282.6)
  310. 310

    Arena rating

    Llama 3 Instruct 70B, Arena rating: 1,277 (scale 900.0 to 1526.2), 95% confidence interval 1273.1 to 1280.31,277(1273.1–1280.3)
  311. 311
    CohereDossier

    Arena rating

    Command R+ (08-2024), Arena rating: 1,276 (scale 900.0 to 1526.2), 95% confidence interval 1269.8 to 1283.01,276(1269.8–1283.0)
  312. 312
    OpenAIDossier

    Arena rating

    GPT-4 (0613), Arena rating: 1,276 (scale 900.0 to 1526.2), 95% confidence interval 1272.4 to 1280.51,276(1272.4–1280.5)
  313. 313
    MistralDossier

    Arena rating

    Mistral Small 3, Arena rating: 1,274 (scale 900.0 to 1526.2), 95% confidence interval 1268.5 to 1280.41,274(1268.5–1280.4)
  314. 314

    Arena rating

    GLM-4 (0520), Arena rating: 1,274 (scale 900.0 to 1526.2), 95% confidence interval 1266.6 to 1280.51,274(1266.6–1280.5)
  315. 315
    Reka AIDossier

    Arena rating

    Reka Flash (Sep '24), Arena rating: 1,273 (scale 900.0 to 1526.2), 95% confidence interval 1265.6 to 1279.61,273(1265.6–1279.6)
  316. 316
    AlibabaDossier

    Arena rating

    Qwen2.5 Coder 32B Instruct, Arena rating: 1,271 (scale 900.0 to 1526.2), 95% confidence interval 1262.6 to 1278.81,271(1262.6–1278.8)
  317. 317
    CohereDossier

    Arena rating

    Aya Expanse 32B, Arena rating: 1,267 (scale 900.0 to 1526.2), 95% confidence interval 1262.5 to 1272.31,267(1262.5–1272.3)
  318. 318
    GoogleDossier

    Arena rating

    Gemma 2 9B IT, Arena rating: 1,267 (scale 900.0 to 1526.2), 95% confidence interval 1263.3 to 1271.01,267(1263.3–1271.0)
  319. 319
    DeepSeekDossier

    Arena rating

    DeepSeek Coder V2, Arena rating: 1,266 (scale 900.0 to 1526.2), 95% confidence interval 1259.3 to 1271.81,266(1259.3–1271.8)
  320. 320
    AlibabaDossier

    Arena rating

    Qwen2 72B Instruct, Arena rating: 1,262 (scale 900.0 to 1526.2), 95% confidence interval 1257.0 to 1266.91,262(1257.0–1266.9)
  321. 321
    CohereDossier

    Arena rating

    Command R+, Arena rating: 1,262 (scale 900.0 to 1526.2), 95% confidence interval 1257.6 to 1266.21,262(1257.6–1266.2)
  322. 322
    AnthropicDossier

    Arena rating

    Claude 3 Haiku, Arena rating: 1,262 (scale 900.0 to 1526.2), 95% confidence interval 1258.0 to 1265.61,262(1258.0–1265.6)
  323. 323
    AmazonDossier

    Arena rating

    Nova Lite, Arena rating: 1,260 (scale 900.0 to 1526.2), 95% confidence interval 1255.3 to 1265.71,260(1255.3–1265.7)
  324. 324
    GoogleDossier

    Arena rating

    Gemini 1.5 Flash-8B (001), Arena rating: 1,259 (scale 900.0 to 1526.2), 95% confidence interval 1254.6 to 1263.21,259(1254.6–1263.2)
  325. 325
    MicrosoftDossier

    Arena rating

    Phi-4, Arena rating: 1,256 (scale 900.0 to 1526.2), 95% confidence interval 1251.8 to 1261.01,256(1251.8–1261.0)
  326. 326
    Allen Institute for AIDossier

    Arena rating

    OLMo 2 32B Instruct (0325), Arena rating: 1,252 (scale 900.0 to 1526.2), 95% confidence interval 1241.0 to 1262.51,252(1241.0–1262.5)
  327. 327
    CohereDossier

    Arena rating

    Command R (08-2024), Arena rating: 1,251 (scale 900.0 to 1526.2), 95% confidence interval 1244.0 to 1257.21,251(1244.0–1257.2)
  328. 328
    MistralDossier

    Arena rating

    Mistral Large (Feb '24), Arena rating: 1,243 (scale 900.0 to 1526.2), 95% confidence interval 1238.0 to 1247.41,243(1238.0–1247.4)
  329. 329
    AmazonDossier

    Arena rating

    Nova Micro, Arena rating: 1,241 (scale 900.0 to 1526.2), 95% confidence interval 1235.6 to 1246.01,241(1235.6–1246.0)
  330. 330
    AI21 LabsDossier

    Arena rating

    Jamba 1.5 Mini, Arena rating: 1,240 (scale 900.0 to 1526.2), 95% confidence interval 1232.7 to 1247.11,240(1232.7–1247.1)
  331. 331
    MistralDossier

    Arena rating

    Ministral 8B (2410), Arena rating: 1,238 (scale 900.0 to 1526.2), 95% confidence interval 1228.6 to 1246.81,238(1228.6–1246.8)
  332. 332
    GoogleDossier

    dev-api

    Arena rating

    Gemini Pro, Arena rating: 1,237 (scale 900.0 to 1526.2), 95% confidence interval 1229.7 to 1244.41,237(1229.7–1244.4)
  333. 333
    AlibabaDossier

    Arena rating

    Qwen1.5 110B Chat, Arena rating: 1,234 (scale 900.0 to 1526.2), 95% confidence interval 1228.8 to 1239.91,234(1228.8–1239.9)
  334. 334
    Reka AIDossier

    online

    Arena rating

    Reka Flash 21B, Arena rating: 1,234 (scale 900.0 to 1526.2), 95% confidence interval 1226.3 to 1241.11,234(1226.3–1241.1)
  335. 335
    AlibabaDossier

    Arena rating

    Qwen1.5 72B Chat, Arena rating: 1,234 (scale 900.0 to 1526.2), 95% confidence interval 1228.4 to 1239.01,234(1228.4–1239.0)
  336. 336
    TencentDossier

    Arena rating

    Hunyuan Standard 256K, Arena rating: 1,233 (scale 900.0 to 1526.2), 95% confidence interval 1221.5 to 1245.01,233(1221.5–1245.0)
  337. 337
    MistralDossier

    Arena rating

    Mixtral 8x22B Instruct v0.1, Arena rating: 1,230 (scale 900.0 to 1526.2), 95% confidence interval 1225.2 to 1234.31,230(1225.2–1234.3)
  338. 338
    CohereDossier

    Arena rating

    Command R, Arena rating: 1,227 (scale 900.0 to 1526.2), 95% confidence interval 1222.3 to 1231.81,227(1222.3–1231.8)
  339. 339
    Reka AIDossier

    Arena rating

    Reka Flash 21B, Arena rating: 1,227 (scale 900.0 to 1526.2), 95% confidence interval 1221.0 to 1232.91,227(1221.0–1232.9)
  340. 340
    OpenAIDossier

    Arena rating

    GPT-3.5 Turbo (0125), Arena rating: 1,226 (scale 900.0 to 1526.2), 95% confidence interval 1221.1 to 1230.51,226(1221.1–1230.5)
  341. 341
    GoogleDossier

    Arena rating

    Gemini Pro, Arena rating: 1,224 (scale 900.0 to 1526.2), 95% confidence interval 1212.1 to 1235.51,224(1212.1–1235.5)
  342. 342

    Arena rating

    Llama 3 Instruct 8B, Arena rating: 1,224 (scale 900.0 to 1526.2), 95% confidence interval 1220.1 to 1227.51,224(1220.1–1227.5)
  343. 343
    CohereDossier

    Arena rating

    Aya Expanse 8B, Arena rating: 1,223 (scale 900.0 to 1526.2), 95% confidence interval 1216.2 to 1230.21,223(1216.2–1230.2)
  344. 344
    MistralDossier

    Arena rating

    Mistral Medium, Arena rating: 1,223 (scale 900.0 to 1526.2), 95% confidence interval 1217.4 to 1228.31,223(1217.4–1228.3)
  345. 345
    Allen Institute for AIDossier

    Arena rating

    Llama 3.1 Tulu 3 8B, Arena rating: 1,221 (scale 900.0 to 1526.2), 95% confidence interval 1209.9 to 1231.21,221(1209.9–1231.2)
  346. 346
    Hugging FaceDossier

    Arena rating

    Zephyr ORPO 141B A35B v0.1, Arena rating: 1,213 (scale 900.0 to 1526.2), 95% confidence interval 1202.4 to 1224.01,213(1202.4–1224.0)
  347. 347
    01.AIDossier

    Arena rating

    Yi 1.5 34B Chat, Arena rating: 1,213 (scale 900.0 to 1526.2), 95% confidence interval 1207.9 to 1218.01,213(1207.9–1218.0)
  348. 348

    Arena rating

    Llama 3.1 Instruct 8B, Arena rating: 1,211 (scale 900.0 to 1526.2), 95% confidence interval 1207.1 to 1215.51,211(1207.1–1215.5)
  349. 349

    Arena rating

    Granite 3.1 8B Instruct, Arena rating: 1,209 (scale 900.0 to 1526.2), 95% confidence interval 1197.6 to 1219.81,209(1197.6–1219.8)
  350. 350
    OpenAIDossier

    Arena rating

    GPT-3.5 Turbo (1106), Arena rating: 1,204 (scale 900.0 to 1526.2), 95% confidence interval 1195.7 to 1213.21,204(1195.7–1213.2)
  351. 351
    AlibabaDossier

    Arena rating

    Qwen1.5 32B Chat, Arena rating: 1,204 (scale 900.0 to 1526.2), 95% confidence interval 1198.1 to 1210.31,204(1198.1–1210.3)
  352. 352
    GoogleDossier

    Arena rating

    Gemma 2 2B IT, Arena rating: 1,200 (scale 900.0 to 1526.2), 95% confidence interval 1196.1 to 1204.31,200(1196.1–1204.3)
  353. 353
    MicrosoftDossier

    Arena rating

    Phi-3 Medium 4K Instruct, Arena rating: 1,198 (scale 900.0 to 1526.2), 95% confidence interval 1192.9 to 1203.21,198(1192.9–1203.2)
  354. 354
    MistralDossier

    Arena rating

    Mixtral 8x7B Instruct, Arena rating: 1,197 (scale 900.0 to 1526.2), 95% confidence interval 1193.1 to 1201.61,197(1193.1–1201.6)
  355. 355
    DatabricksDossier

    Arena rating

    DBRX Instruct Preview, Arena rating: 1,196 (scale 900.0 to 1526.2), 95% confidence interval 1189.6 to 1201.81,196(1189.6–1201.8)
  356. 356
    AlibabaDossier

    Arena rating

    Qwen1.5 14B Chat, Arena rating: 1,191 (scale 900.0 to 1526.2), 95% confidence interval 1184.1 to 1198.31,191(1184.1–1198.3)
  357. 357
    Shanghai AI LaboratoryDossier

    Arena rating

    InternLM2.5 20B Chat, Arena rating: 1,191 (scale 900.0 to 1526.2), 95% confidence interval 1183.7 to 1198.01,191(1183.7–1198.0)
  358. 358
    DeepSeekDossier

    Arena rating

    DeepSeek LLM 67B Chat, Arena rating: 1,185 (scale 900.0 to 1526.2), 95% confidence interval 1173.7 to 1196.61,185(1173.7–1196.6)
  359. 359
    MicrosoftDossier

    Arena rating

    WizardLM 70B, Arena rating: 1,185 (scale 900.0 to 1526.2), 95% confidence interval 1175.5 to 1194.31,185(1175.5–1194.3)
  360. 360
    01.AIDossier

    Arena rating

    Yi 34B Chat, Arena rating: 1,184 (scale 900.0 to 1526.2), 95% confidence interval 1177.1 to 1190.71,184(1177.1–1190.7)
  361. 361

    Arena rating

    Granite 3.0 8B Instruct, Arena rating: 1,183 (scale 900.0 to 1526.2), 95% confidence interval 1174.7 to 1192.01,183(1174.7–1192.0)
  362. 362
    OpenChatDossier

    Arena rating

    OpenChat 3.5, Arena rating: 1,183 (scale 900.0 to 1526.2), 95% confidence interval 1173.6 to 1192.91,183(1173.6–1192.9)
  363. 363
    OpenChatDossier

    Arena rating

    OpenChat 3.5 (0106), Arena rating: 1,183 (scale 900.0 to 1526.2), 95% confidence interval 1174.9 to 1190.91,183(1174.9–1190.9)
  364. 364
    GoogleDossier

    Arena rating

    Gemma 1.1 7B IT, Arena rating: 1,183 (scale 900.0 to 1526.2), 95% confidence interval 1176.7 to 1188.81,183(1176.7–1188.8)
  365. 365
    SnowflakeDossier

    Arena rating

    Snowflake Arctic Instruct, Arena rating: 1,180 (scale 900.0 to 1526.2), 95% confidence interval 1174.3 to 1186.11,180(1174.3–1186.1)
  366. 366

    Arena rating

    Granite 3.1 2B Instruct, Arena rating: 1,179 (scale 900.0 to 1526.2), 95% confidence interval 1167.8 to 1190.21,179(1167.8–1190.2)
  367. 367
    Allen Institute for AIDossier

    Arena rating

    Tulu 2 DPO 70B, Arena rating: 1,178 (scale 900.0 to 1526.2), 95% confidence interval 1168.0 to 1187.61,178(1168.0–1187.6)
  368. 368
    TekniumDossier

    Arena rating

    OpenHermes 2.5 Mistral 7B, Arena rating: 1,176 (scale 900.0 to 1526.2), 95% confidence interval 1165.8 to 1186.51,176(1165.8–1186.5)
  369. 369
    LMSYSDossier

    Arena rating

    Vicuna 33B, Arena rating: 1,173 (scale 900.0 to 1526.2), 95% confidence interval 1166.9 to 1179.31,173(1166.9–1179.3)
  370. 370
    MicrosoftDossier

    Arena rating

    Phi-3 Small 8K Instruct, Arena rating: 1,171 (scale 900.0 to 1526.2), 95% confidence interval 1165.3 to 1177.11,171(1165.3–1177.1)
  371. 371
    NexusflowDossier

    Arena rating

    Starling LM 7B Beta, Arena rating: 1,171 (scale 900.0 to 1526.2), 95% confidence interval 1163.7 to 1178.51,171(1163.7–1178.5)
  372. 372

    Arena rating

    Llama 2 70B Chat, Arena rating: 1,171 (scale 900.0 to 1526.2), 95% confidence interval 1165.3 to 1176.31,171(1165.3–1176.3)
  373. 373
    UC BerkeleyDossier

    Arena rating

    Starling LM 7B Alpha, Arena rating: 1,168 (scale 900.0 to 1526.2), 95% confidence interval 1159.5 to 1175.61,168(1159.5–1175.6)
  374. 374

    Arena rating

    Llama 3.2 3B Instruct, Arena rating: 1,167 (scale 900.0 to 1526.2), 95% confidence interval 1159.2 to 1174.61,167(1159.2–1174.6)
  375. 375
    Nous ResearchDossier

    Arena rating

    Nous Hermes 2 Mixtral 8x7B DPO, Arena rating: 1,165 (scale 900.0 to 1526.2), 95% confidence interval 1152.6 to 1176.41,165(1152.6–1176.4)
  376. 376

    Arena rating

    Granite 3.0 2B Instruct, Arena rating: 1,157 (scale 900.0 to 1526.2), 95% confidence interval 1148.5 to 1165.21,157(1148.5–1165.2)
  377. 377
    NVIDIADossier

    Arena rating

    Llama 2 70B SteerLM Chat, Arena rating: 1,155 (scale 900.0 to 1526.2), 95% confidence interval 1142.3 to 1167.41,155(1142.3–1167.4)
  378. 378
    AlibabaDossier

    Arena rating

    QwQ 32B Preview, Arena rating: 1,154 (scale 900.0 to 1526.2), 95% confidence interval 1142.8 to 1165.71,154(1142.8–1165.7)
  379. 379
    UpstageDossier

    Arena rating

    SOLAR 10.7B Instruct v1.0, Arena rating: 1,153 (scale 900.0 to 1526.2), 95% confidence interval 1139.5 to 1165.71,153(1139.5–1165.7)
  380. 380
    Cognitive ComputationsDossier

    Arena rating

    Dolphin 2.2.1 Mistral 7B, Arena rating: 1,152 (scale 900.0 to 1526.2), 95% confidence interval 1137.2 to 1167.71,152(1137.2–1167.7)
  381. 381
    MosaicMLDossier

    Arena rating

    MPT 30B Chat, Arena rating: 1,151 (scale 900.0 to 1526.2), 95% confidence interval 1138.7 to 1163.11,151(1138.7–1163.1)
  382. 382
    MicrosoftDossier

    Arena rating

    WizardLM 13B, Arena rating: 1,150 (scale 900.0 to 1526.2), 95% confidence interval 1140.5 to 1158.91,150(1140.5–1158.9)
  383. 383
    MistralDossier

    Arena rating

    Mistral 7B Instruct v0.2, Arena rating: 1,150 (scale 900.0 to 1526.2), 95% confidence interval 1142.9 to 1156.21,150(1142.9–1156.2)
  384. 384
    Technology Innovation InstituteDossier

    Arena rating

    Falcon 180B Chat, Arena rating: 1,148 (scale 900.0 to 1526.2), 95% confidence interval 1131.3 to 1165.31,148(1131.3–1165.3)
  385. 385
    AlibabaDossier

    Arena rating

    Qwen1.5 7B Chat, Arena rating: 1,144 (scale 900.0 to 1526.2), 95% confidence interval 1134.1 to 1153.91,144(1134.1–1153.9)
  386. 386
    MicrosoftDossier

    Arena rating

    Phi-3 Mini 4K Instruct (June 2024), Arena rating: 1,143 (scale 900.0 to 1526.2), 95% confidence interval 1136.8 to 1149.71,143(1136.8–1149.7)
  387. 387
    LMSYSDossier

    Arena rating

    Vicuna 13B, Arena rating: 1,142 (scale 900.0 to 1526.2), 95% confidence interval 1135.1 to 1148.41,142(1135.1–1148.4)
  388. 388

    Arena rating

    Llama 2 13B Chat, Arena rating: 1,142 (scale 900.0 to 1526.2), 95% confidence interval 1134.8 to 1148.31,142(1134.8–1148.3)
  389. 389
    AlibabaDossier

    Arena rating

    Qwen 14B Chat, Arena rating: 1,140 (scale 900.0 to 1526.2), 95% confidence interval 1128.6 to 1150.51,140(1128.6–1150.5)
  390. 390
    GoogleDossier

    Arena rating

    PaLM 2, Arena rating: 1,139 (scale 900.0 to 1526.2), 95% confidence interval 1130.1 to 1148.71,139(1130.1–1148.7)
  391. 391
    GoogleDossier

    Arena rating

    Gemma 7B IT, Arena rating: 1,138 (scale 900.0 to 1526.2), 95% confidence interval 1128.5 to 1147.41,138(1128.5–1147.4)
  392. 392

    Arena rating

    Code Llama 34B Instruct, Arena rating: 1,137 (scale 900.0 to 1526.2), 95% confidence interval 1128.2 to 1145.91,137(1128.2–1145.9)
  393. 393
    Hugging FaceDossier

    Arena rating

    Zephyr 7B Beta, Arena rating: 1,131 (scale 900.0 to 1526.2), 95% confidence interval 1122.2 to 1139.81,131(1122.2–1139.8)
  394. 394
    MicrosoftDossier

    Arena rating

    Phi-3 Mini 128K Instruct, Arena rating: 1,130 (scale 900.0 to 1526.2), 95% confidence interval 1122.6 to 1137.41,130(1122.6–1137.4)
  395. 395
    MicrosoftDossier

    Arena rating

    Phi-3 Mini 4K Instruct, Arena rating: 1,128 (scale 900.0 to 1526.2), 95% confidence interval 1121.8 to 1134.61,128(1121.8–1134.6)
  396. 396
    University of WashingtonDossier

    Arena rating

    Guanaco 33B, Arena rating: 1,128 (scale 900.0 to 1526.2), 95% confidence interval 1115.6 to 1139.71,128(1115.6–1139.7)
  397. 397
    Hugging FaceDossier

    Arena rating

    Zephyr 7B Alpha, Arena rating: 1,127 (scale 900.0 to 1526.2), 95% confidence interval 1111.6 to 1142.61,127(1111.6–1142.6)
  398. 398
    Together AIDossier

    Arena rating

    StripedHyena Nous 7B, Arena rating: 1,122 (scale 900.0 to 1526.2), 95% confidence interval 1110.9 to 1132.71,122(1110.9–1132.7)
  399. 399

    Arena rating

    Code Llama 70B Instruct, Arena rating: 1,119 (scale 900.0 to 1526.2), 95% confidence interval 1101.3 to 1137.41,119(1101.3–1137.4)
  400. 400
    GoogleDossier

    Arena rating

    Gemma 1.1 2B IT, Arena rating: 1,117 (scale 900.0 to 1526.2), 95% confidence interval 1109.2 to 1124.61,117(1109.2–1124.6)
  401. 401
    LMSYSDossier

    Arena rating

    Vicuna 7B, Arena rating: 1,115 (scale 900.0 to 1526.2), 95% confidence interval 1106.4 to 1124.61,115(1106.4–1124.6)
  402. 402
    Hugging FaceDossier

    Arena rating

    SmolLM2 1.7B Instruct, Arena rating: 1,115 (scale 900.0 to 1526.2), 95% confidence interval 1100.5 to 1129.21,115(1100.5–1129.2)
  403. 403

    Arena rating

    Llama 3.2 1B Instruct, Arena rating: 1,111 (scale 900.0 to 1526.2), 95% confidence interval 1103.4 to 1119.11,111(1103.4–1119.1)
  404. 404
    MistralDossier

    Arena rating

    Mistral 7B Instruct, Arena rating: 1,111 (scale 900.0 to 1526.2), 95% confidence interval 1101.3 to 1119.81,111(1101.3–1119.8)
  405. 405

    Arena rating

    Llama 2 7B Chat, Arena rating: 1,108 (scale 900.0 to 1526.2), 95% confidence interval 1101.0 to 1115.11,108(1101.0–1115.1)
  406. 406
    GoogleDossier

    Arena rating

    Gemma 2B IT, Arena rating: 1,094 (scale 900.0 to 1526.2), 95% confidence interval 1082.4 to 1105.31,094(1082.4–1105.3)
  407. 407
    AlibabaDossier

    Arena rating

    Qwen1.5 4B Chat, Arena rating: 1,091 (scale 900.0 to 1526.2), 95% confidence interval 1082.0 to 1100.71,091(1082.0–1100.7)
  408. 408
    Allen Institute for AIDossier

    Arena rating

    OLMo 7B Instruct, Arena rating: 1,074 (scale 900.0 to 1526.2), 95% confidence interval 1062.6 to 1084.91,074(1062.6–1084.9)
  409. 409
    UC BerkeleyDossier

    Arena rating

    Koala 13B, Arena rating: 1,071 (scale 900.0 to 1526.2), 95% confidence interval 1061.3 to 1080.91,071(1061.3–1080.9)
  410. 410
    Stanford UniversityDossier

    Arena rating

    Alpaca 13B, Arena rating: 1,070 (scale 900.0 to 1526.2), 95% confidence interval 1059.1 to 1081.91,070(1059.1–1081.9)
  411. 411
    Nomic AIDossier

    Arena rating

    GPT4All 13B Snoozy, Arena rating: 1,068 (scale 900.0 to 1526.2), 95% confidence interval 1052.8 to 1083.11,068(1052.8–1083.1)
  412. 412
    MosaicMLDossier

    Arena rating

    MPT 7B Chat, Arena rating: 1,064 (scale 900.0 to 1526.2), 95% confidence interval 1051.8 to 1075.61,064(1051.8–1075.6)
  413. 413

    Arena rating

    ChatGLM3 6B, Arena rating: 1,057 (scale 900.0 to 1526.2), 95% confidence interval 1045.1 to 1068.41,057(1045.1–1068.4)
  414. 414

    Arena rating

    RWKV-4 Raven 14B, Arena rating: 1,043 (scale 900.0 to 1526.2), 95% confidence interval 1031.2 to 1053.81,043(1031.2–1053.8)
  415. 415

    Arena rating

    ChatGLM2 6B, Arena rating: 1,025 (scale 900.0 to 1526.2), 95% confidence interval 1011.4 to 1038.51,025(1011.4–1038.5)
  416. 416
    Open-Assistant (LAION)Dossier

    Arena rating

    OpenAssistant Pythia 12B, Arena rating: 1,024 (scale 900.0 to 1526.2), 95% confidence interval 1013.0 to 1034.51,024(1013.0–1034.5)
  417. 417

    Arena rating

    ChatGLM 6B, Arena rating: 996 (scale 900.0 to 1526.2), 95% confidence interval 983.5 to 1008.5996(983.5–1008.5)
  418. 418
    LMSYSDossier

    Arena rating

    FastChat-T5 3B, Arena rating: 993 (scale 900.0 to 1526.2), 95% confidence interval 980.5 to 1005.0993(980.5–1005.0)
  419. 419
    DatabricksDossier

    Arena rating

    Dolly 2.0 12B, Arena rating: 983 (scale 900.0 to 1526.2), 95% confidence interval 969.3 to 996.2983(969.3–996.2)
  420. 420

    Arena rating

    LLaMA 13B, Arena rating: 975 (scale 900.0 to 1526.2), 95% confidence interval 959.8 to 991.1975(959.8–991.1)
  421. 421
    Stability AIDossier

    Arena rating

    StableLM Tuned Alpha 7B, Arena rating: 953 (scale 900.0 to 1526.2), 95% confidence interval 940.8 to 966.0953(940.8–966.0)