Last updated: Oct 14, 2026 59 sources 20 endpoints monitored

#reasoning

234 models tagged #reasoning, ranked by Elo score — models not on the LMArena leaderboard fall back to heat.

234 models
  1. 7 kimi-k3 1489
  2. 12 gpt-5.5 1477
  3. 14 gpt-5 1476
  4. 21 glm-4.5 1469
  5. 22 grok-4.5 1469
  6. 23 glm-5.1 1468
  7. 25 gpt-5.4 1466
  8. 29 kimi-k2.6 1461
  9. 35 hy3 1457
  10. 38 kimi-k2.5 1450
  11. 42 minimax-m3 1444
  12. 45 glm-4.7 1442
  13. 46 grok-4.3 1442
  14. 47 grok-4.7 1442
  15. 49 gpt-5.1 1439
  16. 54 gpt-5.2 1436
  17. 61 mimo-v2.5 1434
  18. 70 glm-4.6 1424
  19. 71 grok-4.6 1424
  20. 87 gpt-5-mini 1390
  21. 89 o4-mini 1390
  22. 90 minimax-m2 1384
  23. 93 glm-4.6v 1377
  24. 94 solar-pro4 1377
  25. 100 minimax-m1 1363
  26. 102 glm-4.5v 1353
  27. 105 o3-mini 1348
  28. 106 qwen3-32b 1347
  29. 107 mercury-2 1346
  30. 108 gpt-5-nano 1337
  31. 111 gpt-oss-20b 1318
  32. 125 glm-5.2 94
  33. 150 glm-5.3 75
  34. 163 glm-5 64
  35. 189 ember-1 41
  36. 190 fugu-max 40
  37. 191 aion-3.5 40
  38. 193 aion-3.0 36
  39. 196 aion-2.0 32
  40. 206 seed-1.6 17
  41. 218 o3-pro 3
  42. 220 o1-pro 1
  43. 225 glm-latest —
  44. 233 o1 —
  45. 234 o3 —