Benchmark Almanac

LMArcade

Claude Fable 5 1,798 leads since July 2026 Claude Opus 5 1,689 leads in Terminus 2 or mini-SWE-agent
Not in model cards

Agents build browser games from the same prompt. Players try two games and vote before learning which models and coding tools made them.

AfterQuery’s leaderboard lists 27 model entries and seven coding tools. Each vote updates game ratings, which feed separate model and tool rankings.

Made by
AfterQuery
Released
July 2026
Kind
Web and app front ends
Level
University
Scored
Ranked by votes
Answers
Code
Tasks
All public
Measures
Preference
Language
English
In and out
Text → Code, Actions
Interaction
Agentic
Environment
Browser
Items from
Live play or votes
Refresh
Live
Run by
A host
In views
Agents, Kept fresh
Strongest model tested
Claude Opus 5 #11 on the capability index
Last model added
July 2026
Scores refreshed
November 20th, 2026
160
Entries on its own leaderboard, 38 models we track

Claude Fable 5 leads with 1,798 among models released by July 2026 Claude Fable 5 leads with 1,798 among models released by July 2026 Claude Opus 5 leads with 1,689 in Terminus 2 or mini-SWE-agent

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran on its own with several settings, we keep its best. Agents other companies built around a model get a pale shade of that model's lab color. Gray marks blends and entries not in our model list, such as fine-tunes and older models. None of these count toward leaders. Agents built around a model, such as aq-gaming + Claude Opus 4.7, are their own entries. We don't count them as the model's score. This view shows only models we track, each at its best. This view keeps only runs in Terminus 2 or mini-SWE-agent, a harness no model’s maker built. A maker’s own agent, such as Claude Code with a Claude model, is left out.

The newest model with a score here came out in July 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

160 entries. Scroll sideways for the rest. Bars start at 1,200.

  1. 1,798 Claude Fable 5
  2. 1,753 aq-gaming + Claude Opus 4.7
  3. 1,752 aq-gaming + GPT-5.5
  4. 1,744 GPT-5.6
  5. 1,744 Claude Opus 5
  6. 1,724 aq-gaming + Claude Fable 5
  7. 1,720 swe-agent + Claude Fable 5
  8. 1,695 GPT-5.5
  9. 1,678 aq-gaming + Claude Opus 4.8
  10. 1,676 aq-gaming + Claude Opus 5
  11. 1,675 swe-agent + Claude Opus 5
  12. 1,673 Claude Opus 4.7
  13. 1,658 Claude Opus 4.8
  14. 1,657 aq-gaming + MiMo-V2.5-Pro
  15. 1,656 Claude Sonnet 4.6
  16. 1,654 opengame + GPT-5.5
  17. 1,652 aq-gaming + Qwen 3.6 Max (Preview)
  18. 1,646 opengame + GPT-5.5 Pro
  19. 1,644 aq-gaming + Claude Sonnet 5
  20. 1,639 aq-gaming + GPT-5.5 Pro
  21. 1,639 swe-agent + Claude Opus 4.8
  22. 1,637 opengame + Claude Fable 5
  23. 1,636 aq-gaming + Claude Sonnet 4.6
  24. 1,624 Gemini 3.5 Flash ( dynamic )
  25. 1,622 Claude Sonnet 5
  26. 1,613 opengame + Claude Opus 4.7
  27. 1,611 opengame + Kimi K3
  28. 1,607 swe-agent + GPT-5.5
  29. 1,605 GPT-5.4
  30. 1,600 swe-agent + Claude Sonnet 5
  31. 1,599 aq-gaming + GPT-5.4
  32. 1,597 aq-gaming + DeepSeek-V4-Flash
  33. 1,597 opengame + Claude Opus 5
  34. 1,596 swe-agent + Claude Opus 4.7
  35. 1,593 opengame + Claude Sonnet 4.6
  36. 1,591 swe-agent + Qwen3.7-Plus
  37. 1,589 Qwen3.7-Plus
  38. 1,586 opengame + Qwen 3.6 Max (Preview)
  39. 1,586 GPT-5.5 Pro
  40. 1,583 opengame + Claude Opus 4.8
  41. 1,582 swe-agent + GPT-5.4
  42. 1,580 opengame + GLM-5.1
  43. 1,579 opengame + GPT-5.4
  44. 1,573 Kimi K3
  45. 1,570 DeepSeek-V4-Pro
  46. 1,563 swe-agent + Claude Sonnet 4.6
  47. 1,557 aq-gaming + GLM-5.1
  48. 1,555 aq-gaming + Kimi K3
  49. 1,554 swe-agent + MiMo-V2.5-Pro
  50. 1,553 Grok 4.20
  51. 1,551 opengame + Claude Sonnet 5
  52. 1,549 opengame + Grok 4.20
  53. 1,546 Gemini 3.1 Pro ( dynamic )
  54. 1,544 aq-gaming + Qwen3.7-Plus
  55. 1,542 GLM-5.1
  56. 1,535 aq-gaming + MiniMax-M3
  57. 1,534 aq-gaming + DeepSeek-V4-Pro
  58. 1,531 Gemini 3 Flash ( dynamic )
  59. 1,528 aq-gaming + Kimi K2.6
  60. 1,527 DeepSeek-V4-Flash
  61. 1,526 swe-agent + Kimi K3
  62. 1,525 swe-agent + Qwen 3.6 Max (Preview)
  63. 1,522 opengame + MiMo-V2.5-Pro
  64. 1,516 Qwen3.5 397B-A17B
  65. 1,513 Kimi K2.6
  66. 1,511 aq-gaming + GPT-5.4 Mini
  67. 1,511 Qwen 3.6 Max (Preview)
  68. 1,508 swe-agent + Grok 4.20
  69. 1,507 aq-gaming + Qwen3.5 397B-A17B
  70. 1,505 Mistral Large 2512
  71. 1,505 opengame + DeepSeek-R1
  72. 1,503 Amazon Nova Premier
  73. 1,502 Command A
  74. 1,502 DeepSeek-R1
  75. 1,500 swe-agent + MiniMax-M3
  76. 1,498 aq-gaming + Qwen3-VL-235B-A22B
  77. 1,498 swe-agent + Kimi K2.7 Code
  78. 1,496 opengame + MiniMax-M2.7
  79. 1,495 ERNIE 4.5 VL
  80. 1,495 opengame + Palmyra X5
  81. 1,495 opengame + Solar Pro 3
  82. 1,495 MiniMax-M2.7
  83. 1,493 Solar Pro 3
  84. 1,490 aq-gaming + Hermes 4 405B
  85. 1,490 MiniMax-M3
  86. 1,489 Jamba Large 1.7
  87. 1,489 Qwen3-VL-235B-A22B
  88. 1,484 opengame + Qwen3.7-Plus
  89. 1,483 Gemma 4 31B IT
  90. 1,477 opengame + Seed2.0 Mini
  91. 1,475 swe-agent + DeepSeek-V4-Pro
  92. 1,472 swe-agent + GLM-5.1
  93. 1,471 aq-gaming + GLM-4.6V
  94. 1,469 aq-gaming + DeepSeek-R1
  95. 1,468 swe-agent + GLM-4.6V
  96. 1,468 opengame + GPT-5.4 Mini
  97. 1,467 Claude Haiku 4.5
  98. 1,467 GLM-4.6V
  99. 1,467 aq-gaming + Seed2.0 Mini
  100. 1,467 MiMo-V2.5-Pro
  101. 1,467 swe-agent + Mistral Medium 3.5
  102. 1,466 opengame + DeepSeek-V4-Pro
  103. 1,464 swe-agent + Hermes 4 405B
  104. 1,462 swe-agent + Palmyra X5
  105. 1,461 opengame + Amazon Nova Premier
  106. 1,461 aq-gaming + Kimi K2.7 Code
  107. 1,460 Hunyuan-A13B-Instruct
  108. 1,459 swe-agent + GPT-5.5 Pro
  109. 1,458 swe-agent + Amazon Nova Premier
  110. 1,458 swe-agent + Seed2.0 Mini
  111. 1,457 Palmyra X5
  112. 1,456 Seed2.0 Mini
  113. 1,455 swe-agent + DeepSeek-V4-Flash
  114. 1,453 aq-gaming + Palmyra X5
  115. 1,453 Hermes 4 405B
  116. 1,449 opengame + Qwen3-VL-235B-A22B
  117. 1,448 swe-agent + Hunyuan-A13B-Instruct
  118. 1,448 aq-gaming + Grok 4.3
  119. 1,447 opengame + Hermes 4 405B
  120. 1,447 aq-gaming + MiniMax-M2.7
  121. 1,445 aq-gaming + Amazon Nova Premier
  122. 1,445 opengame + Kimi K2.7 Code
  123. 1,444 Mistral Medium 3.5
  124. 1,442 aq-gaming + Hunyuan-A13B-Instruct
  125. 1,442 aq-gaming + Solar Pro 3
  126. 1,441 opengame + GLM-4.6V
  127. 1,440 aq-gaming + Grok 4.20
  128. 1,440 swe-agent + Gemma 4 31B IT
  129. 1,439 swe-agent + DeepSeek-R1
  130. 1,439 swe-agent + Qwen3-VL-235B-A22B
  131. 1,439 swe-agent + Solar Pro 3
  132. 1,436 Kimi K2.7 Code
  133. 1,435 opengame + Hunyuan-A13B-Instruct
  134. 1,433 swe-agent + GPT-5.4 Mini
  135. 1,433 swe-agent + MiniMax-M2.7
  136. 1,427 opengame + Kimi K2.6
  137. 1,426 swe-agent + Kimi K2.6
  138. 1,424 opengame + MiniMax-M3
  139. 1,416 opengame + Mistral Medium 3.5
  140. 1,412 GPT-5.4 Mini
  141. 1,400 opengame + Grok 4.3
  142. 1,390 Grok 4.3
  143. 1,389 aq-gaming + Mistral Medium 3.5
  144. 1,387 swe-agent + Qwen3.5 397B-A17B
  145. 1,384 opengame + Claude Haiku 4.5
  146. 1,377 opengame + Qwen3.5 397B-A17B
  147. 1,377 aq-gaming + Gemma 4 31B IT
  148. 1,367 swe-agent + Claude Haiku 4.5
  149. 1,364 opengame + DeepSeek-V4-Flash
  150. 1,361 aq-gaming + Claude Haiku 4.5
  151. 1,357 DeepSeek-V3.2
  152. 1,348 swe-agent + Grok 4.3
  153. 1,338 opengame + DeepSeek-V3.2
  154. 1,333 opengame + Gemma 4 31B IT
  155. 1,305 aq-gaming + DeepSeek-V3.2
  156. 1,302 swe-agent + DeepSeek-V3.2
  157. 1,302 swe-agent + Nemotron 3 Super
  158. 1,297 Nemotron 3 Super
  159. 1,267 opengame + Nemotron 3 Super
  160. 1,218 aq-gaming + Nemotron 3 Super

Bars start at 1,200.

  1. 1 Claude Fable 5 1,798
  2. 2 aq-gaming + Claude Opus 4.7 1,753
  3. 3 aq-gaming + GPT-5.5 medium 1,752
  4. 4 GPT-5.6 xhigh 1,744
  5. 5 Claude Opus 5 1,744
  6. 6 aq-gaming + Claude Fable 5 1,724
  7. 7 swe-agent + Claude Fable 5 1,720
  8. 8 GPT-5.5 medium 1,695
  9. 9 aq-gaming + Claude Opus 4.8 1,678
  10. 10 aq-gaming + Claude Opus 5 1,676
  11. 11 swe-agent + Claude Opus 5 1,675
  12. 12 Claude Opus 4.7 1,673
  13. 13 Claude Opus 4.8 1,658
  14. 14 aq-gaming + MiMo-V2.5-Pro 1,657
  15. 15 Claude Sonnet 4.6 1,656
  16. 16 opengame + GPT-5.5 medium 1,654
  17. 17 aq-gaming + Qwen 3.6 Max (Preview) 1,652
  18. 18 opengame + GPT-5.5 Pro high 1,646
  19. 19 aq-gaming + Claude Sonnet 5 1,644
  20. 20 aq-gaming + GPT-5.5 Pro high 1,639
  21. 21 swe-agent + Claude Opus 4.8 1,639
  22. 22 opengame + Claude Fable 5 1,637
  23. 23 aq-gaming + Claude Sonnet 4.6 1,636
  24. 24 Gemini 3.5 Flash ( dynamic ) 1,624
  25. 25 Claude Sonnet 5 1,622
  26. 26 opengame + Claude Opus 4.7 1,613
  27. 27 opengame + Kimi K3 1,611
  28. 28 swe-agent + GPT-5.5 medium 1,607
  29. 29 GPT-5.4 medium 1,605
  30. 30 swe-agent + Claude Sonnet 5 1,600
  31. 31 aq-gaming + GPT-5.4 medium 1,599
  32. 32 aq-gaming + DeepSeek-V4-Flash 1,597
  33. 33 opengame + Claude Opus 5 1,597
  34. 34 swe-agent + Claude Opus 4.7 1,596
  35. 35 opengame + Claude Sonnet 4.6 1,593
  36. 36 swe-agent + Qwen3.7-Plus 1,591
  37. 37 Qwen3.7-Plus 1,589
  38. 38 opengame + Qwen 3.6 Max (Preview) 1,586
  39. 39 GPT-5.5 Pro high 1,586
  40. 40 opengame + Claude Opus 4.8 1,583
  41. 41 swe-agent + GPT-5.4 medium 1,582
  42. 42 opengame + GLM-5.1 1,580
  43. 43 opengame + GPT-5.4 medium 1,579
  44. 44 Kimi K3 1,573
  45. 45 DeepSeek-V4-Pro 1,570
  46. 46 swe-agent + Claude Sonnet 4.6 1,563
  47. 47 aq-gaming + GLM-5.1 1,557
  48. 48 aq-gaming + Kimi K3 1,555
  49. 49 swe-agent + MiMo-V2.5-Pro 1,554
  50. 50 Grok 4.20 1,553
  51. 51 opengame + Claude Sonnet 5 1,551
  52. 52 opengame + Grok 4.20 1,549
  53. 53 Gemini 3.1 Pro ( dynamic ) 1,546
  54. 54 aq-gaming + Qwen3.7-Plus 1,544
  55. 55 GLM-5.1 1,542
  56. 56 aq-gaming + MiniMax-M3 1,535
  57. 57 aq-gaming + DeepSeek-V4-Pro 1,534
  58. 58 Gemini 3 Flash ( dynamic ) 1,531
  59. 59 aq-gaming + Kimi K2.6 1,528
  60. 60 DeepSeek-V4-Flash 1,527
  61. 61 swe-agent + Kimi K3 1,526
  62. 62 swe-agent + Qwen 3.6 Max (Preview) 1,525
  63. 63 opengame + MiMo-V2.5-Pro 1,522
  64. 64 Qwen3.5 397B-A17B 1,516
  65. 65 Kimi K2.6 1,513
  66. 66 aq-gaming + GPT-5.4 Mini medium 1,511
  67. 67 Qwen 3.6 Max (Preview) 1,511
  68. 68 swe-agent + Grok 4.20 1,508
  69. 69 aq-gaming + Qwen3.5 397B-A17B 1,507
  70. 70 Mistral Large 2512 1,505
  71. 71 opengame + DeepSeek-R1 May 2025 1,505
  72. 72 Amazon Nova Premier 1,503
  73. 73 Command A 1,502
  74. 74 DeepSeek-R1 May 2025 1,502
  75. 75 swe-agent + MiniMax-M3 1,500
  76. 76 aq-gaming + Qwen3-VL-235B-A22B 1,498
  77. 77 swe-agent + Kimi K2.7 Code 1,498
  78. 78 opengame + MiniMax-M2.7 1,496
  79. 79 ERNIE 4.5 VL 1,495
  80. 80 opengame + Palmyra X5 1,495
  81. 81 opengame + Solar Pro 3 1,495
  82. 82 MiniMax-M2.7 1,495
  83. 83 Solar Pro 3 1,493
  84. 84 aq-gaming + Hermes 4 405B 1,490
  85. 85 MiniMax-M3 1,490
  86. 86 Jamba Large 1.7 1,489
  87. 87 Qwen3-VL-235B-A22B 1,489
  88. 88 opengame + Qwen3.7-Plus 1,484
  89. 89 Gemma 4 31B IT 1,483
  90. 90 opengame + Seed2.0 Mini 1,477
  91. 91 swe-agent + DeepSeek-V4-Pro 1,475
  92. 92 swe-agent + GLM-5.1 1,472
  93. 93 aq-gaming + GLM-4.6V 1,471
  94. 94 aq-gaming + DeepSeek-R1 May 2025 1,469
  95. 95 swe-agent + GLM-4.6V 1,468
  96. 96 opengame + GPT-5.4 Mini medium 1,468
  97. 97 Claude Haiku 4.5 1,467
  98. 98 GLM-4.6V 1,467
  99. 99 aq-gaming + Seed2.0 Mini 1,467
  100. 100 MiMo-V2.5-Pro 1,467
  101. 101 swe-agent + Mistral Medium 3.5 1,467
  102. 102 opengame + DeepSeek-V4-Pro 1,466
  103. 103 swe-agent + Hermes 4 405B 1,464
  104. 104 swe-agent + Palmyra X5 1,462
  105. 105 opengame + Amazon Nova Premier 1,461
  106. 106 aq-gaming + Kimi K2.7 Code 1,461
  107. 107 Hunyuan-A13B-Instruct 1,460
  108. 108 swe-agent + GPT-5.5 Pro high 1,459
  109. 109 swe-agent + Amazon Nova Premier 1,458
  110. 110 swe-agent + Seed2.0 Mini 1,458
  111. 111 Palmyra X5 1,457
  112. 112 Seed2.0 Mini 1,456
  113. 113 swe-agent + DeepSeek-V4-Flash 1,455
  114. 114 aq-gaming + Palmyra X5 1,453
  115. 115 Hermes 4 405B 1,453
  116. 116 opengame + Qwen3-VL-235B-A22B 1,449
  117. 117 swe-agent + Hunyuan-A13B-Instruct 1,448
  118. 118 aq-gaming + Grok 4.3 1,448
  119. 119 opengame + Hermes 4 405B 1,447
  120. 120 aq-gaming + MiniMax-M2.7 1,447
  121. 121 aq-gaming + Amazon Nova Premier 1,445
  122. 122 opengame + Kimi K2.7 Code 1,445
  123. 123 Mistral Medium 3.5 1,444
  124. 124 aq-gaming + Hunyuan-A13B-Instruct 1,442
  125. 125 aq-gaming + Solar Pro 3 1,442
  126. 126 opengame + GLM-4.6V 1,441
  127. 127 aq-gaming + Grok 4.20 1,440
  128. 128 swe-agent + Gemma 4 31B IT 1,440
  129. 129 swe-agent + DeepSeek-R1 May 2025 1,439
  130. 130 swe-agent + Qwen3-VL-235B-A22B 1,439
  131. 131 swe-agent + Solar Pro 3 1,439
  132. 132 Kimi K2.7 Code 1,436
  133. 133 opengame + Hunyuan-A13B-Instruct 1,435
  134. 134 swe-agent + GPT-5.4 Mini medium 1,433
  135. 135 swe-agent + MiniMax-M2.7 1,433
  136. 136 opengame + Kimi K2.6 1,427
  137. 137 swe-agent + Kimi K2.6 1,426
  138. 138 opengame + MiniMax-M3 1,424
  139. 139 opengame + Mistral Medium 3.5 1,416
  140. 140 GPT-5.4 Mini medium 1,412
  141. 141 opengame + Grok 4.3 1,400
  142. 142 Grok 4.3 1,390
  143. 143 aq-gaming + Mistral Medium 3.5 1,389
  144. 144 swe-agent + Qwen3.5 397B-A17B 1,387
  145. 145 opengame + Claude Haiku 4.5 1,384
  146. 146 opengame + Qwen3.5 397B-A17B 1,377
  147. 147 aq-gaming + Gemma 4 31B IT 1,377
  148. 148 swe-agent + Claude Haiku 4.5 1,367
  149. 149 opengame + DeepSeek-V4-Flash 1,364
  150. 150 aq-gaming + Claude Haiku 4.5 1,361
  151. 151 DeepSeek-V3.2 1,357
  152. 152 swe-agent + Grok 4.3 1,348
  153. 153 opengame + DeepSeek-V3.2 1,338
  154. 154 opengame + Gemma 4 31B IT 1,333
  155. 155 aq-gaming + DeepSeek-V3.2 1,305
  156. 156 swe-agent + DeepSeek-V3.2 1,302
  157. 157 swe-agent + Nemotron 3 Super 1,302
  158. 158 Nemotron 3 Super 1,297
  159. 159 opengame + Nemotron 3 Super 1,267
  160. 160 aq-gaming + Nemotron 3 Super 1,218

Source: Official LMArcade leaderboard

Source: Official LMArcade leaderboard

Source: Official LMArcade leaderboard

#EntryScoreSettingAgent byDateSource
1 Claude Fable 5 1,798 claude-code Anthropic Jun 2026 Official leaderboard
2 aq-gaming + Claude Opus 4.7 1,753 AfterQuery Apr 2026 Official leaderboard
3 aq-gaming + GPT-5.5 1,752 medium AfterQuery Apr 2026 Official leaderboard
4 GPT-5.6 1,744 xhigh, codex OpenAI Official leaderboard
5 Claude Opus 5 1,744 claude-code Anthropic Jul 2026 Official leaderboard
6 aq-gaming + Claude Fable 5 1,724 AfterQuery Jun 2026 Official leaderboard
7 swe-agent + Claude Fable 5 1,720 Princeton Jun 2026 Official leaderboard
8 GPT-5.5 1,695 medium, codex OpenAI Apr 2026 Official leaderboard
9 aq-gaming + Claude Opus 4.8 1,678 AfterQuery May 2026 Official leaderboard
10 aq-gaming + Claude Opus 5 1,676 AfterQuery Jul 2026 Official leaderboard
11 swe-agent + Claude Opus 5 1,675 Princeton Jul 2026 Official leaderboard
12 Claude Opus 4.7 1,673 mini-swe-agent Princeton Apr 2026 Official leaderboard
13 Claude Opus 4.8 1,658 mini-swe-agent Princeton May 2026 Official leaderboard
14 aq-gaming + MiMo-V2.5-Pro 1,657 AfterQuery Apr 2026 Official leaderboard
15 Claude Sonnet 4.6 1,656 terminus-2 Stanford Feb 2026 Official leaderboard
16 opengame + GPT-5.5 1,654 medium CUHK MMLab Apr 2026 Official leaderboard
17 aq-gaming + Qwen 3.6 Max (Preview) 1,652 AfterQuery Apr 2026 Official leaderboard
18 opengame + GPT-5.5 Pro 1,646 high CUHK MMLab Apr 2026 Official leaderboard
19 aq-gaming + Claude Sonnet 5 1,644 AfterQuery Jun 2026 Official leaderboard
20 aq-gaming + GPT-5.5 Pro 1,639 high AfterQuery Apr 2026 Official leaderboard
21 swe-agent + Claude Opus 4.8 1,639 Princeton May 2026 Official leaderboard
22 opengame + Claude Fable 5 1,637 CUHK MMLab Jun 2026 Official leaderboard
23 aq-gaming + Claude Sonnet 4.6 1,636 AfterQuery Feb 2026 Official leaderboard
24 Gemini 3.5 Flash ( dynamic ) 1,624 swe-agent Princeton Official leaderboard
25 Claude Sonnet 5 1,622 mini-swe-agent Princeton Jun 2026 Official leaderboard
26 opengame + Claude Opus 4.7 1,613 CUHK MMLab Apr 2026 Official leaderboard
27 opengame + Kimi K3 1,611 AfterQuery Jul 2026 Official leaderboard
28 swe-agent + GPT-5.5 1,607 medium Princeton Apr 2026 Official leaderboard
29 GPT-5.4 1,605 medium, terminus-2 Stanford Mar 2026 Official leaderboard
30 swe-agent + Claude Sonnet 5 1,600 Princeton Jun 2026 Official leaderboard
31 aq-gaming + GPT-5.4 1,599 medium AfterQuery Mar 2026 Official leaderboard
32 aq-gaming + DeepSeek-V4-Flash 1,597 AfterQuery Apr 2026 Official leaderboard
33 opengame + Claude Opus 5 1,597 AfterQuery Jul 2026 Official leaderboard
34 swe-agent + Claude Opus 4.7 1,596 Princeton Apr 2026 Official leaderboard
35 opengame + Claude Sonnet 4.6 1,593 CUHK MMLab Feb 2026 Official leaderboard
36 swe-agent + Qwen3.7-Plus 1,591 Princeton Jun 2026 Official leaderboard
37 Qwen3.7-Plus 1,589 mini-swe-agent Princeton Jun 2026 Official leaderboard
38 opengame + Qwen 3.6 Max (Preview) 1,586 CUHK MMLab Apr 2026 Official leaderboard
39 GPT-5.5 Pro 1,586 high, terminus-2 Stanford Apr 2026 Official leaderboard
40 opengame + Claude Opus 4.8 1,583 CUHK MMLab May 2026 Official leaderboard
41 swe-agent + GPT-5.4 1,582 medium Princeton Mar 2026 Official leaderboard
42 opengame + GLM-5.1 1,580 CUHK MMLab Apr 2026 Official leaderboard
43 opengame + GPT-5.4 1,579 medium CUHK MMLab Mar 2026 Official leaderboard
44 Kimi K3 1,573 mini-swe-agent Princeton Jul 2026 Official leaderboard
45 DeepSeek-V4-Pro 1,570 mini-swe-agent Princeton Apr 2026 Official leaderboard
46 swe-agent + Claude Sonnet 4.6 1,563 Princeton Feb 2026 Official leaderboard
47 aq-gaming + GLM-5.1 1,557 AfterQuery Apr 2026 Official leaderboard
48 aq-gaming + Kimi K3 1,555 AfterQuery Jul 2026 Official leaderboard
49 swe-agent + MiMo-V2.5-Pro 1,554 Princeton Apr 2026 Official leaderboard
50 Grok 4.20 1,553 terminus-2 Stanford Feb 2026 Official leaderboard
51 opengame + Claude Sonnet 5 1,551 CUHK MMLab Jun 2026 Official leaderboard
52 opengame + Grok 4.20 1,549 CUHK MMLab Feb 2026 Official leaderboard
53 Gemini 3.1 Pro ( dynamic ) 1,546 swe-agent Princeton Official leaderboard
54 aq-gaming + Qwen3.7-Plus 1,544 AfterQuery Jun 2026 Official leaderboard
55 GLM-5.1 1,542 terminus-2 Stanford Apr 2026 Official leaderboard
56 aq-gaming + MiniMax-M3 1,535 AfterQuery Jun 2026 Official leaderboard
57 aq-gaming + DeepSeek-V4-Pro 1,534 AfterQuery Apr 2026 Official leaderboard
58 Gemini 3 Flash ( dynamic ) 1,531 swe-agent Princeton Official leaderboard
59 aq-gaming + Kimi K2.6 1,528 AfterQuery Apr 2026 Official leaderboard
60 DeepSeek-V4-Flash 1,527 mini-swe-agent Princeton Apr 2026 Official leaderboard
61 swe-agent + Kimi K3 1,526 Princeton Jul 2026 Official leaderboard
62 swe-agent + Qwen 3.6 Max (Preview) 1,525 Princeton Apr 2026 Official leaderboard
63 opengame + MiMo-V2.5-Pro 1,522 CUHK MMLab Apr 2026 Official leaderboard
64 Qwen3.5 397B-A17B 1,516 mini-swe-agent Princeton Feb 2026 Official leaderboard
65 Kimi K2.6 1,513 mini-swe-agent Princeton Apr 2026 Official leaderboard
66 aq-gaming + GPT-5.4 Mini 1,511 medium AfterQuery Mar 2026 Official leaderboard
67 Qwen 3.6 Max (Preview) 1,511 terminus-2 Stanford Apr 2026 Official leaderboard
68 swe-agent + Grok 4.20 1,508 Princeton Feb 2026 Official leaderboard
69 aq-gaming + Qwen3.5 397B-A17B 1,507 AfterQuery Feb 2026 Official leaderboard
70 Mistral Large 2512 1,505 terminus-2 Stanford Official leaderboard
71 opengame + DeepSeek-R1 May 2025 1,505 CUHK MMLab May 2025 Official leaderboard
72 Amazon Nova Premier 1,503 mini-swe-agent Princeton Apr 2025 Official leaderboard
73 Command A 1,502 swe-agent Princeton Official leaderboard
74 DeepSeek-R1 May 2025 1,502 mini-swe-agent Princeton May 2025 Official leaderboard
75 swe-agent + MiniMax-M3 1,500 Princeton Jun 2026 Official leaderboard
76 aq-gaming + Qwen3-VL-235B-A22B 1,498 AfterQuery Sep 2025 Official leaderboard
77 swe-agent + Kimi K2.7 Code 1,498 Princeton Jun 2026 Official leaderboard
78 opengame + MiniMax-M2.7 1,496 CUHK MMLab Mar 2026 Official leaderboard
79 ERNIE 4.5 VL 1,495 aq-gaming AfterQuery Official leaderboard
80 opengame + Palmyra X5 1,495 CUHK MMLab Apr 2025 Official leaderboard
81 opengame + Solar Pro 3 1,495 CUHK MMLab Jan 2026 Official leaderboard
82 MiniMax-M2.7 1,495 terminus-2 Stanford Mar 2026 Official leaderboard
83 Solar Pro 3 1,493 terminus-2 Stanford Jan 2026 Official leaderboard
84 aq-gaming + Hermes 4 405B 1,490 AfterQuery Aug 2025 Official leaderboard
85 MiniMax-M3 1,490 mini-swe-agent Princeton Jun 2026 Official leaderboard
86 Jamba Large 1.7 1,489 swe-agent Princeton Official leaderboard
87 Qwen3-VL-235B-A22B 1,489 mini-swe-agent Princeton Sep 2025 Official leaderboard
88 opengame + Qwen3.7-Plus 1,484 CUHK MMLab Jun 2026 Official leaderboard
89 Gemma 4 31B IT 1,483 mini-swe-agent Princeton Apr 2026 Official leaderboard
90 opengame + Seed2.0 Mini 1,477 CUHK MMLab Feb 2026 Official leaderboard
91 swe-agent + DeepSeek-V4-Pro 1,475 Princeton Apr 2026 Official leaderboard
92 swe-agent + GLM-5.1 1,472 Princeton Apr 2026 Official leaderboard
93 aq-gaming + GLM-4.6V 1,471 AfterQuery Dec 2025 Official leaderboard
94 aq-gaming + DeepSeek-R1 May 2025 1,469 AfterQuery May 2025 Official leaderboard
95 swe-agent + GLM-4.6V 1,468 Princeton Dec 2025 Official leaderboard
96 opengame + GPT-5.4 Mini 1,468 medium CUHK MMLab Mar 2026 Official leaderboard
97 Claude Haiku 4.5 1,467 mini-swe-agent Princeton Oct 2025 Official leaderboard
98 GLM-4.6V 1,467 terminus-2 Stanford Dec 2025 Official leaderboard
99 aq-gaming + Seed2.0 Mini 1,467 AfterQuery Feb 2026 Official leaderboard
100 MiMo-V2.5-Pro 1,467 terminus-2 Stanford Apr 2026 Official leaderboard
101 swe-agent + Mistral Medium 3.5 1,467 Princeton Apr 2026 Official leaderboard
102 opengame + DeepSeek-V4-Pro 1,466 CUHK MMLab Apr 2026 Official leaderboard
103 swe-agent + Hermes 4 405B 1,464 Princeton Aug 2025 Official leaderboard
104 swe-agent + Palmyra X5 1,462 Princeton Apr 2025 Official leaderboard
105 opengame + Amazon Nova Premier 1,461 CUHK MMLab Apr 2025 Official leaderboard
106 aq-gaming + Kimi K2.7 Code 1,461 AfterQuery Jun 2026 Official leaderboard
107 Hunyuan-A13B-Instruct 1,460 mini-swe-agent Princeton Jun 2025 Official leaderboard
108 swe-agent + GPT-5.5 Pro 1,459 high Princeton Apr 2026 Official leaderboard
109 swe-agent + Amazon Nova Premier 1,458 Princeton Apr 2025 Official leaderboard
110 swe-agent + Seed2.0 Mini 1,458 Princeton Feb 2026 Official leaderboard
111 Palmyra X5 1,457 mini-swe-agent Princeton Apr 2025 Official leaderboard
112 Seed2.0 Mini 1,456 mini-swe-agent Princeton Feb 2026 Official leaderboard
113 swe-agent + DeepSeek-V4-Flash 1,455 Princeton Apr 2026 Official leaderboard
114 aq-gaming + Palmyra X5 1,453 AfterQuery Apr 2025 Official leaderboard
115 Hermes 4 405B 1,453 mini-swe-agent Princeton Aug 2025 Official leaderboard
116 opengame + Qwen3-VL-235B-A22B 1,449 CUHK MMLab Sep 2025 Official leaderboard
117 swe-agent + Hunyuan-A13B-Instruct 1,448 Princeton Jun 2025 Official leaderboard
118 aq-gaming + Grok 4.3 1,448 AfterQuery Apr 2026 Official leaderboard
119 opengame + Hermes 4 405B 1,447 CUHK MMLab Aug 2025 Official leaderboard
120 aq-gaming + MiniMax-M2.7 1,447 AfterQuery Mar 2026 Official leaderboard
121 aq-gaming + Amazon Nova Premier 1,445 AfterQuery Apr 2025 Official leaderboard
122 opengame + Kimi K2.7 Code 1,445 CUHK MMLab Jun 2026 Official leaderboard
123 Mistral Medium 3.5 1,444 mini-swe-agent Princeton Apr 2026 Official leaderboard
124 aq-gaming + Hunyuan-A13B-Instruct 1,442 AfterQuery Jun 2025 Official leaderboard
125 aq-gaming + Solar Pro 3 1,442 AfterQuery Jan 2026 Official leaderboard
126 opengame + GLM-4.6V 1,441 CUHK MMLab Dec 2025 Official leaderboard
127 aq-gaming + Grok 4.20 1,440 AfterQuery Feb 2026 Official leaderboard
128 swe-agent + Gemma 4 31B IT 1,440 Princeton Apr 2026 Official leaderboard
129 swe-agent + DeepSeek-R1 May 2025 1,439 Princeton May 2025 Official leaderboard
130 swe-agent + Qwen3-VL-235B-A22B 1,439 Princeton Sep 2025 Official leaderboard
131 swe-agent + Solar Pro 3 1,439 Princeton Jan 2026 Official leaderboard
132 Kimi K2.7 Code 1,436 terminus-2 Stanford Jun 2026 Official leaderboard
133 opengame + Hunyuan-A13B-Instruct 1,435 CUHK MMLab Jun 2025 Official leaderboard
134 swe-agent + GPT-5.4 Mini 1,433 medium Princeton Mar 2026 Official leaderboard
135 swe-agent + MiniMax-M2.7 1,433 Princeton Mar 2026 Official leaderboard
136 opengame + Kimi K2.6 1,427 CUHK MMLab Apr 2026 Official leaderboard
137 swe-agent + Kimi K2.6 1,426 Princeton Apr 2026 Official leaderboard
138 opengame + MiniMax-M3 1,424 CUHK MMLab Jun 2026 Official leaderboard
139 opengame + Mistral Medium 3.5 1,416 CUHK MMLab Apr 2026 Official leaderboard
140 GPT-5.4 Mini 1,412 medium, mini-swe-agent Princeton Mar 2026 Official leaderboard
141 opengame + Grok 4.3 1,400 CUHK MMLab Apr 2026 Official leaderboard
142 Grok 4.3 1,390 mini-swe-agent Princeton Apr 2026 Official leaderboard
143 aq-gaming + Mistral Medium 3.5 1,389 AfterQuery Apr 2026 Official leaderboard
144 swe-agent + Qwen3.5 397B-A17B 1,387 Princeton Feb 2026 Official leaderboard
145 opengame + Claude Haiku 4.5 1,384 CUHK MMLab Oct 2025 Official leaderboard
146 opengame + Qwen3.5 397B-A17B 1,377 CUHK MMLab Feb 2026 Official leaderboard
147 aq-gaming + Gemma 4 31B IT 1,377 AfterQuery Apr 2026 Official leaderboard
148 swe-agent + Claude Haiku 4.5 1,367 Princeton Oct 2025 Official leaderboard
149 opengame + DeepSeek-V4-Flash 1,364 CUHK MMLab Apr 2026 Official leaderboard
150 aq-gaming + Claude Haiku 4.5 1,361 AfterQuery Oct 2025 Official leaderboard
151 DeepSeek-V3.2 1,357 terminus-2 Stanford Dec 2025 Official leaderboard
152 swe-agent + Grok 4.3 1,348 Princeton Apr 2026 Official leaderboard
153 opengame + DeepSeek-V3.2 1,338 CUHK MMLab Dec 2025 Official leaderboard
154 opengame + Gemma 4 31B IT 1,333 CUHK MMLab Apr 2026 Official leaderboard
155 aq-gaming + DeepSeek-V3.2 1,305 AfterQuery Dec 2025 Official leaderboard
156 swe-agent + DeepSeek-V3.2 1,302 Princeton Dec 2025 Official leaderboard
157 swe-agent + Nemotron 3 Super 1,302 Princeton Mar 2026 Official leaderboard
158 Nemotron 3 Super 1,297 terminus-2 Stanford Mar 2026 Official leaderboard
159 opengame + Nemotron 3 Super 1,267 CUHK MMLab Mar 2026 Official leaderboard
160 aq-gaming + Nemotron 3 Super 1,218 AfterQuery Mar 2026 Official leaderboard
#ModelScoreSettingAgent byDateSource
1 Claude Fable 5 1,798 claude-code Anthropic Jun 2026 Official leaderboard
2 Claude Opus 5 1,744 claude-code Anthropic Jul 2026 Official leaderboard
3 GPT-5.5 1,695 medium, codex OpenAI Apr 2026 Official leaderboard
4 Claude Opus 4.7 1,673 mini-swe-agent Princeton Apr 2026 Official leaderboard
5 Claude Opus 4.8 1,658 mini-swe-agent Princeton May 2026 Official leaderboard
6 Claude Sonnet 4.6 1,656 terminus-2 Stanford Feb 2026 Official leaderboard
7 Claude Sonnet 5 1,622 mini-swe-agent Princeton Jun 2026 Official leaderboard
8 GPT-5.4 1,605 medium, terminus-2 Stanford Mar 2026 Official leaderboard
9 Qwen3.7-Plus 1,589 mini-swe-agent Princeton Jun 2026 Official leaderboard
10 GPT-5.5 Pro 1,586 high, terminus-2 Stanford Apr 2026 Official leaderboard
11 Kimi K3 1,573 mini-swe-agent Princeton Jul 2026 Official leaderboard
12 DeepSeek-V4-Pro 1,570 mini-swe-agent Princeton Apr 2026 Official leaderboard
13 Grok 4.20 1,553 terminus-2 Stanford Feb 2026 Official leaderboard
14 GLM-5.1 1,542 terminus-2 Stanford Apr 2026 Official leaderboard
15 DeepSeek-V4-Flash 1,527 mini-swe-agent Princeton Apr 2026 Official leaderboard
16 Qwen3.5 397B-A17B 1,516 mini-swe-agent Princeton Feb 2026 Official leaderboard
17 Kimi K2.6 1,513 mini-swe-agent Princeton Apr 2026 Official leaderboard
18 Qwen 3.6 Max (Preview) 1,511 terminus-2 Stanford Apr 2026 Official leaderboard
19 Amazon Nova Premier 1,503 mini-swe-agent Princeton Apr 2025 Official leaderboard
20 DeepSeek-R1 May 2025 1,502 mini-swe-agent Princeton May 2025 Official leaderboard
21 MiniMax-M2.7 1,495 terminus-2 Stanford Mar 2026 Official leaderboard
22 Solar Pro 3 1,493 terminus-2 Stanford Jan 2026 Official leaderboard
23 MiniMax-M3 1,490 mini-swe-agent Princeton Jun 2026 Official leaderboard
24 Qwen3-VL-235B-A22B 1,489 mini-swe-agent Princeton Sep 2025 Official leaderboard
25 Gemma 4 31B IT 1,483 mini-swe-agent Princeton Apr 2026 Official leaderboard
26 Claude Haiku 4.5 1,467 mini-swe-agent Princeton Oct 2025 Official leaderboard
27 GLM-4.6V 1,467 terminus-2 Stanford Dec 2025 Official leaderboard
28 MiMo-V2.5-Pro 1,467 terminus-2 Stanford Apr 2026 Official leaderboard
29 Hunyuan-A13B-Instruct 1,460 mini-swe-agent Princeton Jun 2025 Official leaderboard
30 Palmyra X5 1,457 mini-swe-agent Princeton Apr 2025 Official leaderboard
31 Seed2.0 Mini 1,456 mini-swe-agent Princeton Feb 2026 Official leaderboard
32 Hermes 4 405B 1,453 mini-swe-agent Princeton Aug 2025 Official leaderboard
33 Mistral Medium 3.5 1,444 mini-swe-agent Princeton Apr 2026 Official leaderboard
34 Kimi K2.7 Code 1,436 terminus-2 Stanford Jun 2026 Official leaderboard
35 GPT-5.4 Mini 1,412 medium, mini-swe-agent Princeton Mar 2026 Official leaderboard
36 Grok 4.3 1,390 mini-swe-agent Princeton Apr 2026 Official leaderboard
37 DeepSeek-V3.2 1,357 terminus-2 Stanford Dec 2025 Official leaderboard
38 Nemotron 3 Super 1,297 terminus-2 Stanford Mar 2026 Official leaderboard
#ModelScoreSettingAgent byDateSource
1 Claude Opus 5 1,689 terminus-2 Stanford Jul 2026 Official leaderboard
2 Claude Fable 5 1,687 terminus-2 Stanford Jun 2026 Official leaderboard
3 GPT-5.5 1,686 medium, terminus-2 Stanford Apr 2026 Official leaderboard
4 Claude Opus 4.7 1,673 mini-swe-agent Princeton Apr 2026 Official leaderboard
5 Claude Opus 4.8 1,658 mini-swe-agent Princeton May 2026 Official leaderboard
6 Claude Sonnet 4.6 1,656 terminus-2 Stanford Feb 2026 Official leaderboard
7 Claude Sonnet 5 1,622 mini-swe-agent Princeton Jun 2026 Official leaderboard
8 GPT-5.4 1,605 medium, terminus-2 Stanford Mar 2026 Official leaderboard
9 Qwen3.7-Plus 1,589 mini-swe-agent Princeton Jun 2026 Official leaderboard
10 GPT-5.5 Pro 1,586 high, terminus-2 Stanford Apr 2026 Official leaderboard
11 Kimi K3 1,573 mini-swe-agent Princeton Jul 2026 Official leaderboard
12 DeepSeek-V4-Pro 1,570 mini-swe-agent Princeton Apr 2026 Official leaderboard
13 Grok 4.20 1,553 terminus-2 Stanford Feb 2026 Official leaderboard
14 GLM-5.1 1,542 terminus-2 Stanford Apr 2026 Official leaderboard
15 DeepSeek-V4-Flash 1,527 mini-swe-agent Princeton Apr 2026 Official leaderboard
16 Qwen3.5 397B-A17B 1,516 mini-swe-agent Princeton Feb 2026 Official leaderboard
17 Kimi K2.6 1,513 mini-swe-agent Princeton Apr 2026 Official leaderboard
18 Qwen 3.6 Max (Preview) 1,511 terminus-2 Stanford Apr 2026 Official leaderboard
19 Amazon Nova Premier 1,503 mini-swe-agent Princeton Apr 2025 Official leaderboard
20 DeepSeek-R1 May 2025 1,502 mini-swe-agent Princeton May 2025 Official leaderboard
21 MiniMax-M2.7 1,495 terminus-2 Stanford Mar 2026 Official leaderboard
22 Solar Pro 3 1,493 terminus-2 Stanford Jan 2026 Official leaderboard
23 MiniMax-M3 1,490 mini-swe-agent Princeton Jun 2026 Official leaderboard
24 Qwen3-VL-235B-A22B 1,489 mini-swe-agent Princeton Sep 2025 Official leaderboard
25 Gemma 4 31B IT 1,483 mini-swe-agent Princeton Apr 2026 Official leaderboard
26 Claude Haiku 4.5 1,467 mini-swe-agent Princeton Oct 2025 Official leaderboard
27 GLM-4.6V 1,467 terminus-2 Stanford Dec 2025 Official leaderboard
28 MiMo-V2.5-Pro 1,467 terminus-2 Stanford Apr 2026 Official leaderboard
29 Hunyuan-A13B-Instruct 1,460 mini-swe-agent Princeton Jun 2025 Official leaderboard
30 Palmyra X5 1,457 mini-swe-agent Princeton Apr 2025 Official leaderboard
31 Seed2.0 Mini 1,456 mini-swe-agent Princeton Feb 2026 Official leaderboard
32 Hermes 4 405B 1,453 mini-swe-agent Princeton Aug 2025 Official leaderboard
33 Mistral Medium 3.5 1,444 mini-swe-agent Princeton Apr 2026 Official leaderboard
34 Kimi K2.7 Code 1,436 terminus-2 Stanford Jun 2026 Official leaderboard
35 GPT-5.4 Mini 1,412 medium, mini-swe-agent Princeton Mar 2026 Official leaderboard
36 Grok 4.3 1,390 mini-swe-agent Princeton Apr 2026 Official leaderboard
37 DeepSeek-V3.2 1,357 terminus-2 Stanford Dec 2025 Official leaderboard
38 Nemotron 3 Super 1,297 terminus-2 Stanford Mar 2026 Official leaderboard

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard Not tested since Jul 2026

Source: Official LMArcade leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
1,200 1,400 1,600 1,800 2025 2026 1,200 1,400 1,600 1,800 2025 2026

Source: Official LMArcade leaderboard

On the benchmark’s own leaderboard
1,200 1,300 1,400 1,500 1,600 1,700 2025 2026 1,200 1,300 1,400 1,500 1,600 1,700 2025 2026

Source: Official LMArcade leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

All 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (160 entries).
  • Pass Models scored 38 models we track, plus 122 other entries on its leaderboard.
  • Pass Newest score The newest is from July 2026, for Claude Opus 5.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 5 (#11).
  • Note Headroom Scored as an Elo rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on lmarcade.com: 193 rows for models we track and 8 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check