Skip to content
AI.info

Model Tracker

Text Arena leaderboard

The models in the Model Tracker that Arena has measured on Text Arena, ranked by the highest score of each: the name Arena gives it, the setting it ran and the day. Arena’s leaderboard dataset (lmarena-ai/leaderboard-dataset on Hugging Face

The models in the Model Tracker that Arena has measured on Text Arena, ranked by the highest score of each: the name Arena gives it, the setting it ran and the day. Arena’s leaderboard dataset (lmarena-ai/leaderboard-dataset on Hugging Face), CC BY 4.0: ratings from people’s votes between two anonymous models, as Arena publishes them.

  1. 1. Gemini 4 Argon: 1,524 (1,515–1,533) — Gemini 4 Argon (High), High, 15 October 2026
  2. 2. Muse Wonder: 1,509 (1,498–1,519) — Muse Wonder (Max), Max, 15 October 2026
  3. 3. Claude Opus 5.5: 1,508 (1,501–1,515) — Claude Opus 5.5 (High), High, 15 October 2026
  4. 4. Claude Fable 5: 1,504 (1,499–1,508) — Claude Fable 5 (High), High, 15 October 2026
  5. 5. Claude Opus 4.6: 1,503 (1,500–1,506) — Claude Opus 4.6 (High), High, 15 October 2026
  6. 6. Claude Fable 5.1: 1,502 (1,496–1,508) — Claude Fable 5.1 (Max), Max, 15 October 2026
  7. 7. Claude Opus 4.7: 1,501 (1,497–1,505) — Claude Opus 4.7 (High), High, 15 October 2026
  8. 8. Gemini 3.8 Flash: 1,496 (1,492–1,501) — Gemini 3.8 Flash (High), High, 15 October 2026
  9. 9. Claude Opus 5: 1,490 (1,485–1,495) — Claude Opus 5 (Max), Max, 15 October 2026
  10. 10. Muse Spark: 1,489 (1,483–1,495) — Muse Spark, 15 October 2026
  11. 11. Gemini 3.7 Flash: 1,488 (1,483–1,492) — Gemini 3.7 Flash (High), High, 15 October 2026
  12. 12. Gemini 3 Pro: 1,485 (1,482–1,489) — Gemini 3 Pro, 15 October 2026
  13. 13. GPT-5.6 Sol: 1,485 (1,480–1,489) — gpt-5.6-sol-xhigh, xhigh, 15 October 2026
  14. 14. GPT-6.1 Sol: 1,484 (1,476–1,492) — gpt-6.1-sol-max, max, 15 October 2026
  15. 15. Gemini 3.6 Flash: 1,483 (1,478–1,487) — gemini-3.6-flash-high, high, 15 October 2026
  16. 16. Qwen3.8-Max: 1,482 (1,477–1,487) — qwen3.8-max, 15 October 2026
  17. 17. Claude Opus 4.8: 1,481 (1,478–1,485) — claude-opus-4-8-high, high, 15 October 2026
  18. 18. GPT-5.5: 1,481 (1,478–1,485) — gpt-5.5-high, high, 15 October 2026
  19. 19. Gemini 3.5 Flash: 1,477 (1,474–1,481) — gemini-3.5-flash-high, high, 15 October 2026
  20. 20. GPT-6 Astra: 1,477 (1,471–1,484) — gpt-6-astra-max, max, 15 October 2026
  21. 21. GPT-5.4: 1,475 (1,471–1,479) — gpt-5.4-high, high, 15 October 2026
  22. 22. Claude Sonnet 5.5: 1,475 (1,467–1,483) — claude-sonnet-5.5-xhigh, xhigh, 15 October 2026
  23. 23. DeepSeek-V4.1-Flash: 1,475 (1,468–1,481) — deepseek-v4.1-flash-max, max, 15 October 2026
  24. 24. Claude Opus 4.5: 1,474 (1,470–1,477) — claude-opus-4-5-20251101-high-32k, high-32k, 15 October 2026
  25. 25. Gemini 3 Flash: 1,473 (1,468–1,477) — gemini-3-flash, 15 October 2026
  26. 26. Claude Sonnet 4.6: 1,472 (1,469–1,476) — claude-sonnet-4-6, 15 October 2026
  27. 27. GPT-5.6 Terra: 1,466 (1,462–1,471) — gpt-5.6-terra-xhigh, xhigh, 15 October 2026
  28. 28. Grok 4.5: 1,466 (1,461–1,470) — grok-4.5, 15 October 2026
  29. 29. Grok 4.1: 1,465 (1,462–1,468) — grok-4.1-thinking, thinking, 15 October 2026
  30. 30. DeepSeek V4 Pro: 1,465 (1,458–1,471) — deepseek-v4-pro-high-20260813, high, 15 October 2026
  31. 31. Claude Sonnet 5: 1,461 (1,457–1,465) — claude-sonnet-5-high, high, 15 October 2026
  32. 32. GPT-6 Sol: 1,459 (1,453–1,465) — gpt-6-sol-max, max, 15 October 2026
  33. 33. Claude Sonnet 4.5: 1,456 (1,454–1,459) — claude-sonnet-4-5-20250929-high-32k, high-32k, 15 October 2026
  34. 34. GPT-5.1: 1,456 (1,452–1,459) — gpt-5.1-high, high, 15 October 2026
  35. 35. Gemini 3.5 Flash-Lite: 1,455 (1,450–1,459) — gemini-3.5-flash-lite, 15 October 2026
  36. 36. Grok 4.6: 1,454 (1,450–1,459) — grok-4.6-high, high, 15 October 2026
  37. 37. GPT-5.6 Luna: 1,453 (1,448–1,457) — gpt-5.6-luna-xhigh, xhigh, 15 October 2026
  38. 38. Claude Opus 4.1: 1,450 (1,446–1,453) — claude-opus-4-1-20250805-thinking-16k, thinking-16k, 15 October 2026
  39. 39. GPT-5.4 Mini: 1,447 (1,443–1,451) — gpt-5.4-mini-high, high, 15 October 2026
  40. 40. Gemini 2.5 Pro: 1,445 (1,443–1,448) — gemini-2.5-pro, 15 October 2026
  41. 41. GPT-4.5 Preview: 1,445 (1,439–1,450) — gpt-4.5-preview-2025-02-27, 15 October 2026
  42. 42. Grok 4.7: 1,444 (1,438–1,450) — grok-4.7-xhigh, xhigh, 15 October 2026
  43. 43. GPT-6 Luna: 1,443 (1,438–1,447) — gpt-6-luna-max, max, 15 October 2026
  44. 44. Grok 4.3: 1,442 (1,438–1,445) — grok-4.3, 15 October 2026
  45. 45. Qwen3.8-27B: 1,438 (1,433–1,443) — qwen3.8-27b, 15 October 2026
  46. 46. GPT-5.2: 1,437 (1,434–1,441) — gpt-5.2-high, high, 15 October 2026
  47. 47. DeepSeek-V4-Flash: 1,436 (1,432–1,440) — deepseek-v4-flash, 15 October 2026
  48. 48. GPT-5: 1,435 (1,430–1,439) — gpt-5-high, high, 15 October 2026
  49. 49. Mistral Large 4: 1,432 (1,425–1,439) — mistral-large-4, 15 October 2026
  50. 50. o3: 1,432 (1,428–1,435) — o3-2025-04-16, 15 October 2026
  51. 51. Grok 4.1 Fast: 1,430 (1,427–1,434) — grok-4-1-fast-reasoning, reasoning, 15 October 2026
  52. 52. Mistral Medium 3.5: 1,426 (1,420–1,433) — mistral-medium-3.5, 15 October 2026
  53. 53. GPT-5 Chat: 1,426 (1,422–1,431) — gpt-5-chat, 15 October 2026
  54. 54. DeepSeek-V3.2-Exp: 1,425 (1,419–1,432) — deepseek-v3.2-exp-thinking, thinking, 15 October 2026
  55. 55. DeepSeek-V3.1: 1,417 (1,411–1,423) — deepseek-v3.1, 15 October 2026
  56. 56. Qwen3.5-122B-A10B: 1,416 (1,412–1,421) — qwen3.5-122b-a10b, 15 October 2026
  57. 57. DeepSeek-V3.1-Terminus: 1,416 (1,406–1,426) — deepseek-v3.1-terminus-thinking, thinking, 15 October 2026
  58. 58. GPT-4.1: 1,415 (1,411–1,418) — gpt-4.1-2025-04-14, 15 October 2026
  59. 59. Claude Haiku 4.5: 1,414 (1,412–1,417) — claude-haiku-4-5-20251001, 15 October 2026
  60. 60. Mistral Large 3: 1,414 (1,411–1,417) — mistral-large-3, 15 October 2026
  61. 61. Gemini 2.5 Flash: 1,409 (1,407–1,412) — gemini-2.5-flash, 15 October 2026
  62. 62. Qwen3.5-27B: 1,409 (1,404–1,413) — qwen3.5-27b, 15 October 2026
  63. 63. Grok 4 Fast: 1,405 (1,400–1,410) — grok-4-fast-reasoning, reasoning, 15 October 2026
  64. 64. o1: 1,402 (1,398–1,407) — o1-2024-12-17, 15 October 2026
  65. 65. Claude Sonnet 4: 1,402 (1,397–1,406) — claude-sonnet-4-20250514-thinking-32k, thinking-32k, 15 October 2026
  66. 66. GPT-5.4 nano: 1,401 (1,397–1,405) — gpt-5.4-nano-high, high, 15 October 2026
  67. 67. DeepSeek-R1: 1,398 (1,393–1,403) — deepseek-r1, 15 October 2026
  68. 68. Qwen3.5-35B-A3B: 1,394 (1,390–1,398) — qwen3.5-35b-a3b, 15 October 2026
  69. 69. o4-mini: 1,391 (1,387–1,395) — o4-mini-2025-04-16, 15 October 2026
  70. 70. GPT-5 Mini: 1,390 (1,385–1,395) — gpt-5-mini-high, high, 15 October 2026
  71. 71. o1-preview: 1,389 (1,384–1,394) — o1-preview, 15 October 2026
  72. 72. Claude Sonnet 3.7: 1,388 (1,384–1,393) — claude-3-7-sonnet-20250219-thinking-32k, thinking-32k, 15 October 2026
  73. 73. GPT-4.1 mini: 1,383 (1,379–1,387) — gpt-4.1-mini-2025-04-14, 15 October 2026
  74. 74. Qwen3-235B-A22B: 1,375 (1,370–1,380) — qwen3-235b-a22b, 15 October 2026
  75. 75. Grok 3 mini: 1,364 (1,358–1,369) — grok-3-mini-high, high, 15 October 2026
  76. 76. o3-mini: 1,364 (1,358–1,369) — o3-mini-high, high, 15 October 2026
  77. 77. DeepSeek-V3: 1,358 (1,354–1,363) — deepseek-v3, 15 October 2026
  78. 78. gpt-oss-120b: 1,352 (1,347–1,356) — gpt-oss-120b, 15 October 2026
  79. 79. Qwen3-32B: 1,347 (1,337–1,356) — qwen3-32b, 15 October 2026
  80. 80. GPT-4o: 1,346 (1,343–1,350) — gpt-4o-2024-05-13, 15 October 2026
  81. 81. GPT-5 nano: 1,338 (1,331–1,345) — gpt-5-nano-high, high, 15 October 2026
  82. 82. o1-mini: 1,337 (1,333–1,340) — o1-mini, 15 October 2026
  83. 83. Grok-2: 1,336 (1,332–1,339) — grok-2-2024-08-13, 15 October 2026
  84. 84. Nova 2 Lite: 1,335 (1,329–1,341) — nova-2-lite, 15 October 2026
  85. 85. Qwen3-30B-A3B: 1,327 (1,322–1,331) — qwen3-30b-a3b, 15 October 2026
  86. 86. Claude 3.5 Haiku: 1,325 (1,322–1,328) — claude-3-5-haiku-20241022, 15 October 2026
  87. 87. GPT-4 Turbo: 1,325 (1,321–1,328) — gpt-4-turbo-2024-04-09, 15 October 2026
  88. 88. Claude 3 Opus: 1,322 (1,319–1,325) — claude-3-opus-20240229, 15 October 2026
  89. 89. GPT-4.1 nano: 1,322 (1,315–1,330) — gpt-4.1-nano-2025-04-14, 15 October 2026
  90. 90. GPT-4o mini: 1,318 (1,314–1,321) — gpt-4o-mini-2024-07-18, 15 October 2026
  91. 91. gpt-oss-20b: 1,317 (1,311–1,324) — gpt-oss-20b, 15 October 2026
  92. 92. DeepSeek-V2.5: 1,307 (1,302–1,312) — deepseek-v2.5, 15 October 2026
  93. 93. Claude 3 Sonnet: 1,281 (1,277–1,285) — claude-3-sonnet-20240229, 15 October 2026
  94. 94. DeepSeek-Coder-V2: 1,265 (1,259–1,272) — deepseek-coder-v2, 15 October 2026
  95. 95. Claude 3 Haiku: 1,262 (1,258–1,266) — claude-3-haiku-20240307, 15 October 2026
  96. 96. Command R: 1,227 (1,222–1,232) — command-r, 15 October 2026
  97. 97. Phi-3-small-8k-instruct: 1,171 (1,165–1,177) — phi-3-small-8k-instruct, 15 October 2026
  98. 98. Phi-3-mini-4k-instruct: 1,128 (1,122–1,135) — phi-3-mini-4k-instruct, 15 October 2026

Arena: the benchmark and its data · CC BY 4.0

Other benchmarks

All of the Model Tracker