Benchmark Almanac

Vibe Code Bench

Claude Fable 5.2 93.8% leads since November 2026
Claude Fable 5.2 93.8% leads since November 2026
In use
Version 1.1

The official leaderboard hasn’t scored a model released after Gemini 4 Neon (November 2026). Labs report their own numbers for one newer model: Inkling 2 Mini at 77.9%.

End-to-end web application development from natural-language specs. Built apps are tested against functional requirements.

Official scaffold fork at github.com/vals-ai/VibeCodeBench-Openhands-Scaffold. Results come from the Vals AI leaderboard.

Made by
Vals AI
Released
March 2026
Kind
Building whole projects
Level
Expert
Scored
Graded by a model
Answers
Code
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text → Code
Interaction
Agentic
Environment
Terminal
Run by
A host
In views
Agents
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last in a launch post
November 2026
Last model added
November 2026
Scores refreshed
November 19th, 2026

In the authors’ words

Code generation has emerged as one of AI's highest-impact use cases, yet existing benchmarks measure isolated tasks rather than the complete "zero-to-one" process of building a working application from scratch. We introduce Vibe Code Bench, a benchmark of 100 web application specifications (50 private validation, 50 held-out test) with 964 browser-based workflows comprising 10,131 substeps, evaluated against deployed applications by an autonomous browser agent. Across 16 frontier models, the best achieves 61.8% accuracy on the test split, revealing that reliable end-to-end application development remains a frontier challenge. We identify self-testing during generation as a strong performance predictor (Pearson r=0.72), and show through a completed human alignment study that evaluator selection materially affects outcomes (31.8-93.6% pairwise step-level agreement). Our contributions include (1) a novel benchmark dataset and browser-based evaluation pipeline for end-to-end web application development, (2) a comprehensive evaluation of 16 frontier models with cost, latency, and error analysis, and (3) an evaluator alignment protocol with both cross-model and human annotation results.

Abstract of Vibe Code Bench: Evaluating AI Models on End-to-End Web Application Development (arXiv:2603.04601, CC0)
126
Entries on its own leaderboard, 124 models we track
3
Launch posts list it, from 2 labs
2
System cards report it, from 2 labs

Claude Fable 5.2 leads with 93.8% Claude Fable 5.2 leads with 93.8%

Best score per entry from the benchmark's own leaderboard, read on November 19th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

Claude Fable 5.2 leads with 93.8%, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model. Striped bars are claims, and dotted bars are reruns by Epoch AI or another board.

126 entries. Scroll sideways for the rest.

  1. 93.8 Claude Fable 5.2
  2. 93.7 Gemini 4 Neon
  3. 92.4 Claude Sonnet 5.5
  4. 91.9 Gemini 4 Argon
  5. 90.4 Claude Haiku 5.5
  6. 90.3 Claude Fable 5
  7. 90.3 Claude Opus 5.5
  8. 90.3 Claude Fable 5.1
  9. 89.9 GPT-6.2 Astra
  10. 89.6 GPT-6 Astra
  11. 89.1 Gemini 4 Carbon
  12. 88.9 GPT-6.1 Sol
  13. 88.4 Claude Opus 5
  14. 87.8 GPT-6 Sol
  15. 87.8 Gemini 4 Hydrogen
  16. 87.7 DeepSeek V4.1 Flash 1017
  17. 87.6 GPT-6.2 Sol
  18. 86.7 Qwen4-Max
  19. 86.2 Grok 4.7
  20. 86.0 Muse Wonder Full
  21. 85.9 Muse Spark 1.3
  22. 85.2 MiMo-V2.6 Pro
  23. 85.0 Kimi K3
  24. 84.9 Kimi K3.1 Pro
  25. 84.7 DeepSeek V4.1 Flash
  26. 84.0 GPT-6.1 Terra
  27. 83.9 Ember-1
  28. 83.7 Muse Ember
  29. 83.1 Muse Wonder
  30. 82.7 Claude Opus 4.8
  31. 82.3 DeepSeek V4 Pro 0813
  32. 81.7 GPT-6 Luna
  33. 81.6 Mistral Large 4
  34. 81.5 Inkling 2 Small
  35. 81.3 Claude Sonnet 5
  36. 81.0 Tencent Hy4
  37. 80.5 GPT-5.6 Sol
  38. 79.1 Muse Spark 1.2
  39. 79.0 MiMo-V2.6 Flash
  40. 78.7 Gemini 3.8 Flash
  41. 78.4 Mistral Large 4 Preview
  42. 78.1 GLM-5.3
  43. 77.5 Tencent Hy4 preview
  44. 77.1 Qwen3.8 Max (0902)
  45. 77.1 GPT-5.6 Luna
  46. 76.2 Grok 4.6
  47. 76.2 Muse Spark Open
  48. 74.8 Hy4 Flash
  49. 74.7 DeepSeek V4 Flash 0731
  50. 74.6 GPT-5.6 Terra
  51. 73.9 MiniMax-M3.1-Flash
  52. 72.2 Muse Spark 1.1
  53. 71.0 Claude Opus 4.7
  54. 70.4 Step 5 Preview
  55. 70.4 Gemini 3.7 Flash
  56. 69.8 GPT-5.5
  57. 69.0 Grok 4.5
  58. 67.4 GPT-5.4
  59. 66.4 Beam 1.0
  60. 64.8 Qwen 3.8 27B
  61. 64.7 Qwen 3.8 Max
  62. 64.0 Gemini 3.6 Flash
  63. 64.0 GLM-5.2
  64. 61.8 GPT-5.3 Codex
  65. 60.9 Mercury 3
  66. 57.6 Claude Opus 4.6
  67. 55.8 Claude Sonnet 4.6
  68. 53.5 GPT-5.2
  69. 49.9 DeepSeek-V4-Pro
  70. 49.6 Composer 2.5
  71. 48.7 Gemini 3.5 Flash
  72. 48.0 GPT-5.4 Mini
  73. 47.7 Qwen3.7-Max
  74. 47.6 MiniMax-M3
  75. 47.2 Kimi K2.7 Code
  76. 46.4 Qwen3.7-Plus
  77. 42.2 MiMo-V2.5
  78. 37.9 GPT-5.2 Codex
  79. 37.9 Kimi K2.6
  80. 37.2 Gemini 3.5 Flash-Lite
  81. 34.1 MiMo-V2.5-Pro
  82. 32.0 Gemini 3.1 Pro
  83. 31.5 GLM-5.1
  84. 30.8 GLM-5.3-Flash
  85. 26.1 GPT-5.4 Nano
  86. 25.6 Qwen 3.6 Plus
  87. 24.6 GPT-5.1
  88. 23.4 GLM-5
  89. 22.6 Claude Sonnet 4.5
  90. 22.2 GPT-5.1-Codex-Max
  91. 20.6 Claude Opus 4.5
  92. 20.2 Gemini 3 Flash
  93. 20.1 GPT-5
  94. 19.7 Muse Spark
  95. 19.4 Grok 4.3
  96. 19.2 Inkling
  97. 19.1 Inkling-Small
  98. 17.5 Kimi K2.5
  99. 15.7 Qwen 3.5 Plus (hosted 397B-A17B)
  100. 14.8 MiniMax-M2.5
  101. 14.3 Gemini 3 Pro
  102. 14.2 GPT-5 mini
  103. 13.3 Grok Build 0.1
  104. 13.1 GPT-5.1-Codex
  105. 11.9 Qwen3.6 27B
  106. 11.9 MiniMax-M2.7
  107. 11.4 Claude Haiku 4.5
  108. 11.0 Laguna M.1
  109. 7.64 Nemotron 3 Ultra
  110. 7.22 SWE-1.6 Fast
  111. 5.21 Laguna XS.2
  112. 5.11 DeepSeek-V3.2
  113. 4.06 Grok 4.20
  114. 3.71 Nemotron 3.5 Super VL
  115. 3.68 Mercury 2.5
  116. 3.51 Qwen3-Max
  117. 3.40 Ling-3.0-flash-Fin
  118. 3.09 GLM-4.6
  119. 2.91 Ling-3.0-flash
  120. 2.89 Mistral Medium 3.5
  121. 1.20 Grok 4.1 Fast
  122. 0.40 Gemini 2.5 Pro
  123. 0.0 Grok 4 Fast
  124. 0.0 Gemini 3.1 Flash-Lite
  125. 0.0 Mistral Small 4
  126. 0.0 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Fable 5.2 93.8%
  2. 2 Gemini 4 Neon 93.7%
  3. 3 Claude Sonnet 5.5 92.4%
  4. 4 Gemini 4 Argon 91.9%
  5. 5 Claude Haiku 5.5 90.4%
  6. 6 Claude Fable 5 90.3%
  7. 7 Claude Opus 5.5 90.3%
  8. 8 Claude Fable 5.1 90.3%
  9. 9 GPT-6.2 Astra 89.9%
  10. 10 GPT-6 Astra 89.6%
  11. 11 Gemini 4 Carbon 89.1%
  12. 12 GPT-6.1 Sol 88.9%
  13. 13 Claude Opus 5 88.4%
  14. 14 GPT-6 Sol 87.8%
  15. 15 Gemini 4 Hydrogen 87.8%
  16. 16 DeepSeek V4.1 Flash 1017 87.7%
  17. 17 GPT-6.2 Sol 87.6%
  18. 18 Qwen4-Max 86.7%
  19. 19 Grok 4.7 86.2%
  20. 20 Muse Wonder Full 86%
  21. 21 Muse Spark 1.3 max 85.9%
  22. 22 MiMo-V2.6 Pro 85.2%
  23. 23 Kimi K3 85%
  24. 24 Kimi K3.1 Pro 84.9%
  25. 25 DeepSeek V4.1 Flash 84.7%
  26. 26 GPT-6.1 Terra 84%
  27. 27 Ember-1 83.9%
  28. 28 Muse Ember 83.7%
  29. 29 Muse Wonder 83.1%
  30. 30 Claude Opus 4.8 82.7%
  31. 31 DeepSeek V4 Pro 0813 82.3%
  32. 32 GPT-6 Luna 81.7%
  33. 33 Mistral Large 4 81.6%
  34. 34 Inkling 2 Small 81.5%
  35. 35 Claude Sonnet 5 81.3%
  36. 36 Tencent Hy4 81%
  37. 37 GPT-5.6 Sol 80.5%
  38. 38 Muse Spark 1.2 79.1%
  39. 39 MiMo-V2.6 Flash 79%
  40. 40 Gemini 3.8 Flash 78.7%
  41. 41 Mistral Large 4 Preview 78.4%
  42. 42 GLM-5.3 78.1%
  43. 43 Tencent Hy4 preview 77.5%
  44. 44 Qwen3.8 Max (0902) 77.1%
  45. 45 GPT-5.6 Luna 77.1%
  46. 46 Grok 4.6 76.2%
  47. 47 Muse Spark Open 76.2%
  48. 48 Hy4 Flash 74.8%
  49. 49 DeepSeek V4 Flash 0731 74.7%
  50. 50 GPT-5.6 Terra 74.6%
  51. 51 MiniMax-M3.1-Flash 73.9%
  52. 52 Muse Spark 1.1 72.2%
  53. 53 Claude Opus 4.7 71%
  54. 54 Step 5 Preview 70.4%
  55. 55 Gemini 3.7 Flash 70.4%
  56. 56 GPT-5.5 69.8%
  57. 57 Grok 4.5 69%
  58. 58 GPT-5.4 xhigh 67.4%
  59. 59 Beam 1.0 66.4%
  60. 60 Qwen 3.8 27B 64.8%
  61. 61 Qwen 3.8 Max 64.7%
  62. 62 Gemini 3.6 Flash 64%
  63. 63 GLM-5.2 64%
  64. 64 GPT-5.3 Codex 61.8%
  65. 65 Mercury 3 60.9%
  66. 66 Claude Opus 4.6 non-thinking 57.6%
  67. 67 Claude Sonnet 4.6 55.8%
  68. 68 GPT-5.2 53.5%
  69. 69 DeepSeek-V4-Pro 49.9%
  70. 70 Composer 2.5 49.6%
  71. 71 Gemini 3.5 Flash 48.7%
  72. 72 GPT-5.4 Mini 48%
  73. 73 Qwen3.7-Max 47.7%
  74. 74 MiniMax-M3 47.6%
  75. 75 Kimi K2.7 Code 47.2%
  76. 76 Qwen3.7-Plus 46.4%
  77. 77 MiMo-V2.5 42.2%
  78. 78 GPT-5.2 Codex 37.9%
  79. 79 Kimi K2.6 37.9%
  80. 80 Gemini 3.5 Flash-Lite 37.2%
  81. 81 MiMo-V2.5-Pro 34.1%
  82. 82 Gemini 3.1 Pro 32%
  83. 83 GLM-5.1 31.5%
  84. 84 GLM-5.3-Flash 30.8%
  85. 85 GPT-5.4 Nano 26.1%
  86. 86 Qwen 3.6 Plus 25.6%
  87. 87 GPT-5.1 24.6%
  88. 88 GLM-5 thinking 23.4%
  89. 89 Claude Sonnet 4.5 thinking 22.6%
  90. 90 GPT-5.1-Codex-Max 22.2%
  91. 91 Claude Opus 4.5 thinking 20.6%
  92. 92 Gemini 3 Flash 20.2%
  93. 93 GPT-5 20.1%
  94. 94 Muse Spark 19.7%
  95. 95 Grok 4.3 19.4%
  96. 96 Inkling 19.2%
  97. 97 Inkling-Small 19.1%
  98. 98 Kimi K2.5 thinking 17.5%
  99. 99 Qwen 3.5 Plus (hosted 397B-A17B) thinking 15.7%
  100. 100 MiniMax-M2.5 14.8%
  101. 101 Gemini 3 Pro 14.3%
  102. 102 GPT-5 mini 14.2%
  103. 103 Grok Build 0.1 13.3%
  104. 104 GPT-5.1-Codex 13.1%
  105. 105 Qwen3.6 27B 11.9%
  106. 106 MiniMax-M2.7 11.9%
  107. 107 Claude Haiku 4.5 thinking 11.4%
  108. 108 Laguna M.1 11%
  109. 109 Nemotron 3 Ultra 7.64%
  110. 110 SWE-1.6 Fast 7.22%
  111. 111 Laguna XS.2 5.21%
  112. 112 DeepSeek-V3.2 thinking 5.11%
  113. 113 Grok 4.20 reasoning 4.06%
  114. 114 Nemotron 3.5 Super VL 3.71%
  115. 115 Mercury 2.5 3.68%
  116. 116 Qwen3-Max 3.51%
  117. 117 Ling-3.0-flash-Fin 3.40%
  118. 118 GLM-4.6 3.09%
  119. 119 Ling-3.0-flash 2.91%
  120. 120 Mistral Medium 3.5 2.89%
  121. 121 Grok 4.1 Fast reasoning 1.20%
  122. 122 Gemini 2.5 Pro Jun 2025 0.40%
  123. 123 Grok 4 Fast reasoning 0%
  124. 124 Gemini 3.1 Flash-Lite 0%
  125. 125 Mistral Small 4 0%
  126. 126 Nemotron 3.5 Lightning 30B-A3B 0%

Source: Official Vibe Code Bench leaderboard

124 models. Scroll sideways for the rest.

  1. 93.8 Claude Fable 5.2
  2. 93.7 Gemini 4 Neon
  3. 92.4 Claude Sonnet 5.5
  4. 91.9 Gemini 4 Argon
  5. 90.4 Claude Haiku 5.5
  6. 90.3 Claude Fable 5
  7. 90.3 Claude Opus 5.5
  8. 90.3 Claude Fable 5.1
  9. 89.9 GPT-6.2 Astra
  10. 89.6 GPT-6 Astra
  11. 89.1 Gemini 4 Carbon
  12. 88.9 GPT-6.1 Sol
  13. 88.4 Claude Opus 5
  14. 87.8 GPT-6 Sol
  15. 87.8 Gemini 4 Hydrogen
  16. 87.7 DeepSeek V4.1 Flash 1017
  17. 87.6 GPT-6.2 Sol
  18. 86.7 Qwen4-Max
  19. 86.2 Grok 4.7
  20. 86.0 Muse Wonder Full
  21. 85.9 Muse Spark 1.3
  22. 85.2 MiMo-V2.6 Pro
  23. 85.0 Kimi K3
  24. 84.9 Kimi K3.1 Pro
  25. 84.7 DeepSeek V4.1 Flash
  26. 84.0 GPT-6.1 Terra
  27. 83.9 Ember-1
  28. 83.7 Muse Ember
  29. 83.1 Muse Wonder
  30. 82.7 Claude Opus 4.8
  31. 82.3 DeepSeek V4 Pro 0813
  32. 81.7 GPT-6 Luna
  33. 81.6 Mistral Large 4
  34. 81.5 Inkling 2 Small
  35. 81.3 Claude Sonnet 5
  36. 81.0 Tencent Hy4
  37. 80.5 GPT-5.6 Sol
  38. 79.1 Muse Spark 1.2
  39. 79.0 MiMo-V2.6 Flash
  40. 78.7 Gemini 3.8 Flash
  41. 78.4 Mistral Large 4 Preview
  42. 78.1 GLM-5.3
  43. 77.5 Tencent Hy4 preview
  44. 77.1 Qwen3.8 Max (0902)
  45. 77.1 GPT-5.6 Luna
  46. 76.2 Grok 4.6
  47. 76.2 Muse Spark Open
  48. 74.7 DeepSeek V4 Flash 0731
  49. 74.6 GPT-5.6 Terra
  50. 73.9 MiniMax-M3.1-Flash
  51. 72.2 Muse Spark 1.1
  52. 71.0 Claude Opus 4.7
  53. 70.4 Step 5 Preview
  54. 70.4 Gemini 3.7 Flash
  55. 69.8 GPT-5.5
  56. 69.0 Grok 4.5
  57. 67.4 GPT-5.4
  58. 66.4 Beam 1.0
  59. 64.8 Qwen 3.8 27B
  60. 64.7 Qwen 3.8 Max
  61. 64.0 Gemini 3.6 Flash
  62. 64.0 GLM-5.2
  63. 61.8 GPT-5.3 Codex
  64. 60.9 Mercury 3
  65. 57.6 Claude Opus 4.6
  66. 55.8 Claude Sonnet 4.6
  67. 53.5 GPT-5.2
  68. 49.9 DeepSeek-V4-Pro
  69. 49.6 Composer 2.5
  70. 48.7 Gemini 3.5 Flash
  71. 48.0 GPT-5.4 Mini
  72. 47.7 Qwen3.7-Max
  73. 47.6 MiniMax-M3
  74. 47.2 Kimi K2.7 Code
  75. 46.4 Qwen3.7-Plus
  76. 42.2 MiMo-V2.5
  77. 37.9 GPT-5.2 Codex
  78. 37.9 Kimi K2.6
  79. 37.2 Gemini 3.5 Flash-Lite
  80. 34.1 MiMo-V2.5-Pro
  81. 32.0 Gemini 3.1 Pro
  82. 31.5 GLM-5.1
  83. 30.8 GLM-5.3-Flash
  84. 26.1 GPT-5.4 Nano
  85. 25.6 Qwen 3.6 Plus
  86. 24.6 GPT-5.1
  87. 23.4 GLM-5
  88. 22.6 Claude Sonnet 4.5
  89. 22.2 GPT-5.1-Codex-Max
  90. 20.6 Claude Opus 4.5
  91. 20.2 Gemini 3 Flash
  92. 20.1 GPT-5
  93. 19.7 Muse Spark
  94. 19.4 Grok 4.3
  95. 19.2 Inkling
  96. 19.1 Inkling-Small
  97. 17.5 Kimi K2.5
  98. 15.7 Qwen 3.5 Plus (hosted 397B-A17B)
  99. 14.8 MiniMax-M2.5
  100. 14.3 Gemini 3 Pro
  101. 14.2 GPT-5 mini
  102. 13.3 Grok Build 0.1
  103. 13.1 GPT-5.1-Codex
  104. 11.9 Qwen3.6 27B
  105. 11.9 MiniMax-M2.7
  106. 11.4 Claude Haiku 4.5
  107. 11.0 Laguna M.1
  108. 7.64 Nemotron 3 Ultra
  109. 5.21 Laguna XS.2
  110. 5.11 DeepSeek-V3.2
  111. 4.06 Grok 4.20
  112. 3.71 Nemotron 3.5 Super VL
  113. 3.68 Mercury 2.5
  114. 3.51 Qwen3-Max
  115. 3.40 Ling-3.0-flash-Fin
  116. 3.09 GLM-4.6
  117. 2.91 Ling-3.0-flash
  118. 2.89 Mistral Medium 3.5
  119. 1.20 Grok 4.1 Fast
  120. 0.40 Gemini 2.5 Pro
  121. 0.0 Grok 4 Fast
  122. 0.0 Gemini 3.1 Flash-Lite
  123. 0.0 Mistral Small 4
  124. 0.0 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Fable 5.2 93.8%
  2. 2 Gemini 4 Neon 93.7%
  3. 3 Claude Sonnet 5.5 92.4%
  4. 4 Gemini 4 Argon 91.9%
  5. 5 Claude Haiku 5.5 90.4%
  6. 6 Claude Fable 5 90.3%
  7. 7 Claude Opus 5.5 90.3%
  8. 8 Claude Fable 5.1 90.3%
  9. 9 GPT-6.2 Astra 89.9%
  10. 10 GPT-6 Astra 89.6%
  11. 11 Gemini 4 Carbon 89.1%
  12. 12 GPT-6.1 Sol 88.9%
  13. 13 Claude Opus 5 88.4%
  14. 14 GPT-6 Sol 87.8%
  15. 15 Gemini 4 Hydrogen 87.8%
  16. 16 DeepSeek V4.1 Flash 1017 87.7%
  17. 17 GPT-6.2 Sol 87.6%
  18. 18 Qwen4-Max 86.7%
  19. 19 Grok 4.7 86.2%
  20. 20 Muse Wonder Full 86%
  21. 21 Muse Spark 1.3 max 85.9%
  22. 22 MiMo-V2.6 Pro 85.2%
  23. 23 Kimi K3 85%
  24. 24 Kimi K3.1 Pro 84.9%
  25. 25 DeepSeek V4.1 Flash 84.7%
  26. 26 GPT-6.1 Terra 84%
  27. 27 Ember-1 83.9%
  28. 28 Muse Ember 83.7%
  29. 29 Muse Wonder 83.1%
  30. 30 Claude Opus 4.8 82.7%
  31. 31 DeepSeek V4 Pro 0813 82.3%
  32. 32 GPT-6 Luna 81.7%
  33. 33 Mistral Large 4 81.6%
  34. 34 Inkling 2 Small 81.5%
  35. 35 Claude Sonnet 5 81.3%
  36. 36 Tencent Hy4 81%
  37. 37 GPT-5.6 Sol 80.5%
  38. 38 Muse Spark 1.2 79.1%
  39. 39 MiMo-V2.6 Flash 79%
  40. 40 Gemini 3.8 Flash 78.7%
  41. 41 Mistral Large 4 Preview 78.4%
  42. 42 GLM-5.3 78.1%
  43. 43 Tencent Hy4 preview 77.5%
  44. 44 Qwen3.8 Max (0902) 77.1%
  45. 45 GPT-5.6 Luna 77.1%
  46. 46 Grok 4.6 76.2%
  47. 47 Muse Spark Open 76.2%
  48. 48 DeepSeek V4 Flash 0731 74.7%
  49. 49 GPT-5.6 Terra 74.6%
  50. 50 MiniMax-M3.1-Flash 73.9%
  51. 51 Muse Spark 1.1 72.2%
  52. 52 Claude Opus 4.7 71%
  53. 53 Step 5 Preview 70.4%
  54. 54 Gemini 3.7 Flash 70.4%
  55. 55 GPT-5.5 69.8%
  56. 56 Grok 4.5 69%
  57. 57 GPT-5.4 xhigh 67.4%
  58. 58 Beam 1.0 66.4%
  59. 59 Qwen 3.8 27B 64.8%
  60. 60 Qwen 3.8 Max 64.7%
  61. 61 Gemini 3.6 Flash 64%
  62. 62 GLM-5.2 64%
  63. 63 GPT-5.3 Codex 61.8%
  64. 64 Mercury 3 60.9%
  65. 65 Claude Opus 4.6 non-thinking 57.6%
  66. 66 Claude Sonnet 4.6 55.8%
  67. 67 GPT-5.2 53.5%
  68. 68 DeepSeek-V4-Pro 49.9%
  69. 69 Composer 2.5 49.6%
  70. 70 Gemini 3.5 Flash 48.7%
  71. 71 GPT-5.4 Mini 48%
  72. 72 Qwen3.7-Max 47.7%
  73. 73 MiniMax-M3 47.6%
  74. 74 Kimi K2.7 Code 47.2%
  75. 75 Qwen3.7-Plus 46.4%
  76. 76 MiMo-V2.5 42.2%
  77. 77 GPT-5.2 Codex 37.9%
  78. 78 Kimi K2.6 37.9%
  79. 79 Gemini 3.5 Flash-Lite 37.2%
  80. 80 MiMo-V2.5-Pro 34.1%
  81. 81 Gemini 3.1 Pro 32%
  82. 82 GLM-5.1 31.5%
  83. 83 GLM-5.3-Flash 30.8%
  84. 84 GPT-5.4 Nano 26.1%
  85. 85 Qwen 3.6 Plus 25.6%
  86. 86 GPT-5.1 24.6%
  87. 87 GLM-5 thinking 23.4%
  88. 88 Claude Sonnet 4.5 thinking 22.6%
  89. 89 GPT-5.1-Codex-Max 22.2%
  90. 90 Claude Opus 4.5 thinking 20.6%
  91. 91 Gemini 3 Flash 20.2%
  92. 92 GPT-5 20.1%
  93. 93 Muse Spark 19.7%
  94. 94 Grok 4.3 19.4%
  95. 95 Inkling 19.2%
  96. 96 Inkling-Small 19.1%
  97. 97 Kimi K2.5 thinking 17.5%
  98. 98 Qwen 3.5 Plus (hosted 397B-A17B) thinking 15.7%
  99. 99 MiniMax-M2.5 14.8%
  100. 100 Gemini 3 Pro 14.3%
  101. 101 GPT-5 mini 14.2%
  102. 102 Grok Build 0.1 13.3%
  103. 103 GPT-5.1-Codex 13.1%
  104. 104 Qwen3.6 27B 11.9%
  105. 105 MiniMax-M2.7 11.9%
  106. 106 Claude Haiku 4.5 thinking 11.4%
  107. 107 Laguna M.1 11%
  108. 108 Nemotron 3 Ultra 7.64%
  109. 109 Laguna XS.2 5.21%
  110. 110 DeepSeek-V3.2 thinking 5.11%
  111. 111 Grok 4.20 reasoning 4.06%
  112. 112 Nemotron 3.5 Super VL 3.71%
  113. 113 Mercury 2.5 3.68%
  114. 114 Qwen3-Max 3.51%
  115. 115 Ling-3.0-flash-Fin 3.40%
  116. 116 GLM-4.6 3.09%
  117. 117 Ling-3.0-flash 2.91%
  118. 118 Mistral Medium 3.5 2.89%
  119. 119 Grok 4.1 Fast reasoning 1.20%
  120. 120 Gemini 2.5 Pro Jun 2025 0.40%
  121. 121 Grok 4 Fast reasoning 0%
  122. 122 Gemini 3.1 Flash-Lite 0%
  123. 123 Mistral Small 4 0%
  124. 124 Nemotron 3.5 Lightning 30B-A3B 0%

Source: Official Vibe Code Bench leaderboard

125 models. Scroll sideways for the rest.

  1. 93.8 Claude Fable 5.2
  2. 93.7 Gemini 4 Neon
  3. 92.4 Claude Sonnet 5.5
  4. 91.9 Gemini 4 Argon
  5. 90.4 Claude Haiku 5.5
  6. 90.3 Claude Fable 5
  7. 90.3 Claude Opus 5.5
  8. 90.3 Claude Fable 5.1
  9. 89.9 GPT-6.2 Astra
  10. 89.6 GPT-6 Astra
  11. 89.1 Gemini 4 Carbon
  12. 88.9 GPT-6.1 Sol
  13. 88.4 Claude Opus 5
  14. 87.8 GPT-6 Sol
  15. 87.8 Gemini 4 Hydrogen
  16. 87.7 DeepSeek V4.1 Flash 1017
  17. 87.6 GPT-6.2 Sol
  18. 86.7 Qwen4-Max
  19. 86.2 Grok 4.7
  20. 86.0 Muse Wonder Full
  21. 85.9 Muse Spark 1.3
  22. 85.2 MiMo-V2.6 Pro
  23. 85.0 Kimi K3
  24. 84.9 Kimi K3.1 Pro
  25. 84.7 DeepSeek V4.1 Flash
  26. 84.0 GPT-6.1 Terra
  27. 83.9 Ember-1
  28. 83.7 Muse Ember
  29. 83.1 Muse Wonder
  30. 82.7 Claude Opus 4.8
  31. 82.3 DeepSeek V4 Pro 0813
  32. 81.7 GPT-6 Luna
  33. 81.6 Inkling 2 Small
  34. 81.6 Mistral Large 4
  35. 81.3 Claude Sonnet 5
  36. 81.0 Tencent Hy4
  37. 80.5 GPT-5.6 Sol
  38. 79.1 Muse Spark 1.2
  39. 79.0 MiMo-V2.6 Flash
  40. 78.7 Gemini 3.8 Flash
  41. 78.4 Mistral Large 4 Preview
  42. 78.1 GLM-5.3
  43. 77.9 Inkling 2 Mini
  44. 77.5 Tencent Hy4 preview
  45. 77.1 Qwen3.8 Max (0902)
  46. 77.1 GPT-5.6 Luna
  47. 76.2 Grok 4.6
  48. 76.2 Muse Spark Open
  49. 74.7 DeepSeek V4 Flash 0731
  50. 74.6 GPT-5.6 Terra
  51. 73.9 MiniMax-M3.1-Flash
  52. 72.2 Muse Spark 1.1
  53. 71.0 Claude Opus 4.7
  54. 70.4 Step 5 Preview
  55. 70.4 Gemini 3.7 Flash
  56. 69.8 GPT-5.5
  57. 69.0 Grok 4.5
  58. 67.4 GPT-5.4
  59. 66.4 Beam 1.0
  60. 64.8 Qwen 3.8 27B
  61. 64.7 Qwen 3.8 Max
  62. 64.0 Gemini 3.6 Flash
  63. 64.0 GLM-5.2
  64. 61.8 GPT-5.3 Codex
  65. 60.9 Mercury 3
  66. 57.6 Claude Opus 4.6
  67. 55.8 Claude Sonnet 4.6
  68. 53.5 GPT-5.2
  69. 49.9 DeepSeek-V4-Pro
  70. 49.6 Composer 2.5
  71. 48.7 Gemini 3.5 Flash
  72. 48.0 GPT-5.4 Mini
  73. 47.7 Qwen3.7-Max
  74. 47.6 MiniMax-M3
  75. 47.2 Kimi K2.7 Code
  76. 46.4 Qwen3.7-Plus
  77. 42.2 MiMo-V2.5
  78. 37.9 GPT-5.2 Codex
  79. 37.9 Kimi K2.6
  80. 37.2 Gemini 3.5 Flash-Lite
  81. 34.1 MiMo-V2.5-Pro
  82. 32.0 Gemini 3.1 Pro
  83. 31.5 GLM-5.1
  84. 30.8 GLM-5.3-Flash
  85. 26.1 GPT-5.4 Nano
  86. 25.6 Qwen 3.6 Plus
  87. 24.6 GPT-5.1
  88. 23.4 GLM-5
  89. 22.6 Claude Sonnet 4.5
  90. 22.2 GPT-5.1-Codex-Max
  91. 20.6 Claude Opus 4.5
  92. 20.2 Gemini 3 Flash
  93. 20.1 GPT-5
  94. 19.7 Muse Spark
  95. 19.4 Grok 4.3
  96. 19.2 Inkling
  97. 19.1 Inkling-Small
  98. 17.5 Kimi K2.5
  99. 15.7 Qwen 3.5 Plus (hosted 397B-A17B)
  100. 14.8 MiniMax-M2.5
  101. 14.3 Gemini 3 Pro
  102. 14.2 GPT-5 mini
  103. 13.3 Grok Build 0.1
  104. 13.1 GPT-5.1-Codex
  105. 11.9 Qwen3.6 27B
  106. 11.9 MiniMax-M2.7
  107. 11.4 Claude Haiku 4.5
  108. 11.0 Laguna M.1
  109. 7.64 Nemotron 3 Ultra
  110. 5.21 Laguna XS.2
  111. 5.11 DeepSeek-V3.2
  112. 4.06 Grok 4.20
  113. 3.71 Nemotron 3.5 Super VL
  114. 3.68 Mercury 2.5
  115. 3.51 Qwen3-Max
  116. 3.40 Ling-3.0-flash-Fin
  117. 3.09 GLM-4.6
  118. 2.91 Ling-3.0-flash
  119. 2.89 Mistral Medium 3.5
  120. 1.20 Grok 4.1 Fast
  121. 0.40 Gemini 2.5 Pro
  122. 0.0 Grok 4 Fast
  123. 0.0 Gemini 3.1 Flash-Lite
  124. 0.0 Mistral Small 4
  125. 0.0 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Fable 5.2 93.8%
  2. 2 Gemini 4 Neon 93.7%
  3. 3 Claude Sonnet 5.5 92.4%
  4. 4 Gemini 4 Argon 91.9%
  5. 5 Claude Haiku 5.5 90.4%
  6. 6 Claude Fable 5 90.3%
  7. 7 Claude Opus 5.5 90.3%
  8. 8 Claude Fable 5.1 90.3%
  9. 9 GPT-6.2 Astra 89.9%
  10. 10 GPT-6 Astra 89.6%
  11. 11 Gemini 4 Carbon 89.1%
  12. 12 GPT-6.1 Sol 88.9%
  13. 13 Claude Opus 5 88.4%
  14. 14 GPT-6 Sol 87.8%
  15. 15 Gemini 4 Hydrogen 87.8%
  16. 16 DeepSeek V4.1 Flash 1017 87.7%
  17. 17 GPT-6.2 Sol 87.6%
  18. 18 Qwen4-Max 86.7%
  19. 19 Grok 4.7 86.2%
  20. 20 Muse Wonder Full 86%
  21. 21 Muse Spark 1.3 max 85.9%
  22. 22 MiMo-V2.6 Pro 85.2%
  23. 23 Kimi K3 85%
  24. 24 Kimi K3.1 Pro 84.9%
  25. 25 DeepSeek V4.1 Flash 84.7%
  26. 26 GPT-6.1 Terra 84%
  27. 27 Ember-1 83.9%
  28. 28 Muse Ember 83.7%
  29. 29 Muse Wonder 83.1%
  30. 30 Claude Opus 4.8 82.7%
  31. 31 DeepSeek V4 Pro 0813 82.3%
  32. 32 GPT-6 Luna 81.7%
  33. 33 Inkling 2 Small 81.6%
  34. 34 Mistral Large 4 81.6%
  35. 35 Claude Sonnet 5 81.3%
  36. 36 Tencent Hy4 81%
  37. 37 GPT-5.6 Sol 80.5%
  38. 38 Muse Spark 1.2 79.1%
  39. 39 MiMo-V2.6 Flash 79%
  40. 40 Gemini 3.8 Flash 78.7%
  41. 41 Mistral Large 4 Preview 78.4%
  42. 42 GLM-5.3 78.1%
  43. 43 Inkling 2 Mini 77.9%
  44. 44 Tencent Hy4 preview 77.5%
  45. 45 Qwen3.8 Max (0902) 77.1%
  46. 46 GPT-5.6 Luna 77.1%
  47. 47 Grok 4.6 76.2%
  48. 48 Muse Spark Open 76.2%
  49. 49 DeepSeek V4 Flash 0731 74.7%
  50. 50 GPT-5.6 Terra 74.6%
  51. 51 MiniMax-M3.1-Flash 73.9%
  52. 52 Muse Spark 1.1 72.2%
  53. 53 Claude Opus 4.7 71%
  54. 54 Step 5 Preview 70.4%
  55. 55 Gemini 3.7 Flash 70.4%
  56. 56 GPT-5.5 69.8%
  57. 57 Grok 4.5 69%
  58. 58 GPT-5.4 xhigh 67.4%
  59. 59 Beam 1.0 66.4%
  60. 60 Qwen 3.8 27B 64.8%
  61. 61 Qwen 3.8 Max 64.7%
  62. 62 Gemini 3.6 Flash 64%
  63. 63 GLM-5.2 64%
  64. 64 GPT-5.3 Codex 61.8%
  65. 65 Mercury 3 60.9%
  66. 66 Claude Opus 4.6 non-thinking 57.6%
  67. 67 Claude Sonnet 4.6 55.8%
  68. 68 GPT-5.2 53.5%
  69. 69 DeepSeek-V4-Pro 49.9%
  70. 70 Composer 2.5 49.6%
  71. 71 Gemini 3.5 Flash 48.7%
  72. 72 GPT-5.4 Mini 48%
  73. 73 Qwen3.7-Max 47.7%
  74. 74 MiniMax-M3 47.6%
  75. 75 Kimi K2.7 Code 47.2%
  76. 76 Qwen3.7-Plus 46.4%
  77. 77 MiMo-V2.5 42.2%
  78. 78 GPT-5.2 Codex 37.9%
  79. 79 Kimi K2.6 37.9%
  80. 80 Gemini 3.5 Flash-Lite 37.2%
  81. 81 MiMo-V2.5-Pro 34.1%
  82. 82 Gemini 3.1 Pro 32%
  83. 83 GLM-5.1 31.5%
  84. 84 GLM-5.3-Flash 30.8%
  85. 85 GPT-5.4 Nano 26.1%
  86. 86 Qwen 3.6 Plus 25.6%
  87. 87 GPT-5.1 24.6%
  88. 88 GLM-5 thinking 23.4%
  89. 89 Claude Sonnet 4.5 thinking 22.6%
  90. 90 GPT-5.1-Codex-Max 22.2%
  91. 91 Claude Opus 4.5 thinking 20.6%
  92. 92 Gemini 3 Flash 20.2%
  93. 93 GPT-5 20.1%
  94. 94 Muse Spark 19.7%
  95. 95 Grok 4.3 19.4%
  96. 96 Inkling 19.2%
  97. 97 Inkling-Small 19.1%
  98. 98 Kimi K2.5 thinking 17.5%
  99. 99 Qwen 3.5 Plus (hosted 397B-A17B) thinking 15.7%
  100. 100 MiniMax-M2.5 14.8%
  101. 101 Gemini 3 Pro 14.3%
  102. 102 GPT-5 mini 14.2%
  103. 103 Grok Build 0.1 13.3%
  104. 104 GPT-5.1-Codex 13.1%
  105. 105 Qwen3.6 27B 11.9%
  106. 106 MiniMax-M2.7 11.9%
  107. 107 Claude Haiku 4.5 thinking 11.4%
  108. 108 Laguna M.1 11%
  109. 109 Nemotron 3 Ultra 7.64%
  110. 110 Laguna XS.2 5.21%
  111. 111 DeepSeek-V3.2 thinking 5.11%
  112. 112 Grok 4.20 reasoning 4.06%
  113. 113 Nemotron 3.5 Super VL 3.71%
  114. 114 Mercury 2.5 3.68%
  115. 115 Qwen3-Max 3.51%
  116. 116 Ling-3.0-flash-Fin 3.40%
  117. 117 GLM-4.6 3.09%
  118. 118 Ling-3.0-flash 2.91%
  119. 119 Mistral Medium 3.5 2.89%
  120. 120 Grok 4.1 Fast reasoning 1.20%
  121. 121 Gemini 2.5 Pro Jun 2025 0.40%
  122. 122 Grok 4 Fast reasoning 0%
  123. 123 Gemini 3.1 Flash-Lite 0%
  124. 124 Mistral Small 4 0%
  125. 125 Nemotron 3.5 Lightning 30B-A3B 0%

Sources: Official Vibe Code Bench leaderboard and labs’ model cards

#EntryScore ± 95% CISettingDateSource
1 Claude Fable 5.2 93.8% ±2.3 OpenHands Nov 2026 Official leaderboard
2 Gemini 4 Neon 93.7% ±2.3 OpenHands Nov 2026 Official leaderboard
3 Claude Sonnet 5.5 92.4% ±2.5 OpenHands Sep 2026 Official leaderboard
4 Gemini 4 Argon 91.9% ±3.7 OpenHands Sep 2026 Official leaderboard
5 Claude Haiku 5.5 90.4% ±3.4 OpenHands Oct 2026 Official leaderboard
6 Claude Fable 5 90.3% ±4.1 OpenHands Jun 2026 Official leaderboard
7 Claude Opus 5.5 90.3% ±3 OpenHands Sep 2026 Official leaderboard
8 Claude Fable 5.1 90.3% ±3.1 OpenHands Sep 2026 Official leaderboard
9 GPT-6.2 Astra 89.9% ±4 OpenHands Nov 2026 Official leaderboard
10 GPT-6 Astra 89.6% ±4.3 OpenHands Sep 2026 Official leaderboard
11 Gemini 4 Carbon 89.1% ±4.5 OpenHands Oct 2026 Official leaderboard
12 GPT-6.1 Sol 88.9% ±3.7 OpenHands Sep 2026 Official leaderboard
13 Claude Opus 5 88.4% ±5.9 OpenHands Jul 2026 Official leaderboard
14 GPT-6 Sol 87.8% ±5 OpenHands Sep 2026 Official leaderboard
15 Gemini 4 Hydrogen 87.8% ±4.5 OpenHands Nov 2026 Official leaderboard
16 DeepSeek V4.1 Flash 1017 87.7% ±5.4 OpenHands Oct 2026 Official leaderboard
17 GPT-6.2 Sol 87.6% ±4.3 OpenHands Nov 2026 Official leaderboard
18 Qwen4-Max 86.7% ±4.5 OpenHands Nov 2026 Official leaderboard
19 Grok 4.7 86.2% ±4.3 OpenHands Sep 2026 Official leaderboard
20 Muse Wonder Full 86% ±3.8 OpenHands Oct 2026 Official leaderboard
21 Muse Spark 1.3 85.9% ±4.9 max, OpenHands Sep 2026 Official leaderboard
22 MiMo-V2.6 Pro 85.2% ±6.6 OpenHands Sep 2026 Official leaderboard
23 Kimi K3 85% ±5.4 OpenHands Jul 2026 Official leaderboard
24 Kimi K3.1 Pro 84.9% ±5.8 OpenHands Oct 2026 Official leaderboard
25 DeepSeek V4.1 Flash 84.7% ±5.7 OpenHands Sep 2026 Official leaderboard
26 GPT-6.1 Terra 84% ±5.6 OpenHands Oct 2026 Official leaderboard
27 Ember-1 83.9% ±6.2 OpenHands Sep 2026 Official leaderboard
28 Muse Ember 83.7% ±5.4 OpenHands Nov 2026 Official leaderboard
29 Muse Wonder 83.1% ±4.9 OpenHands Oct 2026 Official leaderboard
30 Claude Opus 4.8 82.7% ±6 OpenHands May 2026 Official leaderboard
31 DeepSeek V4 Pro 0813 82.3% ±6.2 OpenHands Aug 2026 Official leaderboard
32 GPT-6 Luna 81.7% ±6.6 OpenHands Sep 2026 Official leaderboard
33 Mistral Large 4 81.6% ±6.4 OpenHands Oct 2026 Official leaderboard
34 Inkling 2 Small 81.5% ±6.8 OpenHands Nov 2026 Official leaderboard
35 Claude Sonnet 5 81.3% ±6 OpenHands Jun 2026 Official leaderboard
36 Tencent Hy4 81% ±7 OpenHands Nov 2026 Official leaderboard
37 GPT-5.6 Sol 80.5% ±7.3 OpenHands Jul 2026 Official leaderboard
38 Muse Spark 1.2 79.1% ±6.5 OpenHands Aug 2026 Official leaderboard
39 MiMo-V2.6 Flash 79% ±7.9 OpenHands Sep 2026 Official leaderboard
40 Gemini 3.8 Flash 78.7% ±7.6 OpenHands Sep 2026 Official leaderboard
41 Mistral Large 4 Preview 78.4% ±7 OpenHands Oct 2026 Official leaderboard
42 GLM-5.3 78.1% ±7.7 OpenHands Aug 2026 Official leaderboard
43 Tencent Hy4 preview 77.5% ±7.9 OpenHands Aug 2026 Official leaderboard
44 Qwen3.8 Max (0902) 77.1% ±7.3 OpenHands Sep 2026 Official leaderboard
45 GPT-5.6 Luna 77.1% ±6 OpenHands Jul 2026 Official leaderboard
46 Grok 4.6 76.2% ±7.5 OpenHands Aug 2026 Official leaderboard
47 Muse Spark Open 76.2% ±6.5 OpenHands Oct 2026 Official leaderboard
48 Hy4 Flash 74.8% ±7.6 OpenHands Official leaderboard
49 DeepSeek V4 Flash 0731 74.7% ±6.7 OpenHands Jul 2026 Official leaderboard
50 GPT-5.6 Terra 74.6% ±8.2 OpenHands Jul 2026 Official leaderboard
51 MiniMax-M3.1-Flash 73.9% ±8.3 OpenHands Oct 2026 Official leaderboard
52 Muse Spark 1.1 72.2% ±6.5 OpenHands Jul 2026 Official leaderboard
53 Claude Opus 4.7 71% ±8.8 OpenHands Apr 2026 Official leaderboard
54 Step 5 Preview 70.4% ±8.4 OpenHands Sep 2026 Official leaderboard
55 Gemini 3.7 Flash 70.4% ±9.5 OpenHands Aug 2026 Official leaderboard
56 GPT-5.5 69.8% ±8.9 OpenHands Apr 2026 Official leaderboard
57 Grok 4.5 69% ±8.8 OpenHands Jul 2026 Official leaderboard
58 GPT-5.4 67.4% ±9.5 xhigh, OpenHands Mar 2026 Official leaderboard
59 Beam 1.0 66.4% ±8 OpenHands Oct 2026 Official leaderboard
60 Qwen 3.8 27B 64.8% ±9.5 OpenHands Aug 2026 Official leaderboard
61 Qwen 3.8 Max 64.7% ±10.5 OpenHands Aug 2026 Official leaderboard
62 Gemini 3.6 Flash 64% ±8.4 OpenHands Jul 2026 Official leaderboard
63 GLM-5.2 64% ±9.4 OpenHands Jun 2026 Official leaderboard
64 GPT-5.3 Codex 61.8% ±9.2 OpenHands Feb 2026 Official leaderboard
65 Mercury 3 60.9% ±10.5 OpenHands Oct 2026 Official leaderboard
66 Claude Opus 4.6 57.6% ±8.6 non-thinking, OpenHands Feb 2026 Official leaderboard
67 Claude Sonnet 4.6 55.8% ±9.2 Claude Code Feb 2026 Official leaderboard
68 GPT-5.2 53.5% ±9.9 OpenHands Dec 2025 Official leaderboard
69 DeepSeek-V4-Pro 49.9% ±9.3 OpenHands Apr 2026 Official leaderboard
70 Composer 2.5 49.6% ±9.7 Cursor CLI May 2026 Official leaderboard
71 Gemini 3.5 Flash 48.7% ±9.3 OpenHands May 2026 Official leaderboard
72 GPT-5.4 Mini 48% ±11 OpenHands Mar 2026 Official leaderboard
73 Qwen3.7-Max 47.7% ±9.1 OpenHands May 2026 Official leaderboard
74 MiniMax-M3 47.6% ±10.7 OpenHands Jun 2026 Official leaderboard
75 Kimi K2.7 Code 47.2% ±10.2 OpenHands Jun 2026 Official leaderboard
76 Qwen3.7-Plus 46.4% ±9 OpenHands Jun 2026 Official leaderboard
77 MiMo-V2.5 42.2% ±9 OpenHands Apr 2026 Official leaderboard
78 GPT-5.2 Codex 37.9% ±9 OpenHands Dec 2025 Official leaderboard
79 Kimi K2.6 37.9% ±9.6 OpenHands Apr 2026 Official leaderboard
80 Gemini 3.5 Flash-Lite 37.2% ±9.1 OpenHands Jul 2026 Official leaderboard
81 MiMo-V2.5-Pro 34.1% ±8.9 OpenHands Apr 2026 Official leaderboard
82 Gemini 3.1 Pro 32% ±8.5 OpenHands Feb 2026 Official leaderboard
83 GLM-5.1 31.5% ±8.9 OpenHands Apr 2026 Official leaderboard
84 GLM-5.3-Flash 30.8% ±10.3 OpenHands Aug 2026 Official leaderboard
85 GPT-5.4 Nano 26.1% ±9.9 OpenHands Mar 2026 Official leaderboard
86 Qwen 3.6 Plus 25.6% ±7.7 OpenHands Mar 2026 Official leaderboard
87 GPT-5.1 24.6% ±8.3 OpenHands Nov 2025 Official leaderboard
88 GLM-5 23.4% ±7.9 thinking, OpenHands Feb 2026 Official leaderboard
89 Claude Sonnet 4.5 22.6% ±7.3 thinking, OpenHands Sep 2025 Official leaderboard
90 GPT-5.1-Codex-Max 22.2% ±7.5 OpenHands Nov 2025 Official leaderboard
91 Claude Opus 4.5 20.6% ±6.2 thinking, OpenHands Nov 2025 Official leaderboard
92 Gemini 3 Flash 20.2% ±7.7 OpenHands Dec 2025 Official leaderboard
93 GPT-5 20.1% ±6.7 OpenHands Aug 2025 Official leaderboard
94 Muse Spark 19.7% ±7.8 OpenHands Apr 2026 Official leaderboard
95 Grok 4.3 19.4% ±8.3 OpenHands Apr 2026 Official leaderboard
96 Inkling 19.2% ±7.2 OpenHands Jul 2026 Official leaderboard
97 Inkling-Small 19.1% ±6.7 OpenHands Jul 2026 Official leaderboard
98 Kimi K2.5 17.5% ±6.4 thinking, OpenHands Jan 2026 Official leaderboard
99 Qwen 3.5 Plus (hosted 397B-A17B) 15.7% ±6.2 thinking, OpenHands Feb 2026 Official leaderboard
100 MiniMax-M2.5 14.8% ±5.8 OpenHands Feb 2026 Official leaderboard
101 Gemini 3 Pro 14.3% ±6 OpenHands Nov 2025 Official leaderboard
102 GPT-5 mini 14.2% ±6.9 OpenHands Aug 2025 Official leaderboard
103 Grok Build 0.1 13.3% ±7.5 Grok Build May 2026 Official leaderboard
104 GPT-5.1-Codex 13.1% ±6.2 OpenHands Nov 2025 Official leaderboard
105 Qwen3.6 27B 11.9% ±6.7 OpenHands Apr 2026 Official leaderboard
106 MiniMax-M2.7 11.9% ±5.6 OpenHands Mar 2026 Official leaderboard
107 Claude Haiku 4.5 11.4% ±6.1 thinking, OpenHands Oct 2025 Official leaderboard
108 Laguna M.1 11% ±5.5 OpenHands Apr 2026 Official leaderboard
109 Nemotron 3 Ultra 7.64% ±5.5 OpenHands Jun 2026 Official leaderboard
110 SWE-1.6 Fast 7.22% ±5.2 Devin CLI Official leaderboard
111 Laguna XS.2 5.21% ±4.3 OpenHands Apr 2026 Official leaderboard
112 DeepSeek-V3.2 5.11% ±4.2 thinking, OpenHands Dec 2025 Official leaderboard
113 Grok 4.20 4.06% ±4 reasoning, OpenHands Feb 2026 Official leaderboard
114 Nemotron 3.5 Super VL 3.71% ±3.8 OpenHands Oct 2026 Official leaderboard
115 Mercury 2.5 3.68% ±2.8 OpenHands Sep 2026 Official leaderboard
116 Qwen3-Max 3.51% ±2.5 OpenHands Sep 2025 Official leaderboard
117 Ling-3.0-flash-Fin 3.40% ±2.5 OpenHands Sep 2026 Official leaderboard
118 GLM-4.6 3.09% ±2.3 OpenHands Sep 2025 Official leaderboard
119 Ling-3.0-flash 2.91% ±4.1 OpenHands Aug 2026 Official leaderboard
120 Mistral Medium 3.5 2.89% ±3.4 OpenHands Apr 2026 Official leaderboard
121 Grok 4.1 Fast 1.20% ±2.4 reasoning, OpenHands Nov 2025 Official leaderboard
122 Gemini 2.5 Pro Jun 2025 0.40% ±0.8 OpenHands Jun 2025 Official leaderboard
123 Grok 4 Fast 0% reasoning, OpenHands Sep 2025 Official leaderboard
124 Gemini 3.1 Flash-Lite 0% OpenHands Mar 2026 Official leaderboard
125 Mistral Small 4 0% OpenHands Mar 2026 Official leaderboard
126 Nemotron 3.5 Lightning 30B-A3B 0% OpenHands Aug 2026 Official leaderboard
#ModelScore ± 95% CISettingDateSource
1 Claude Fable 5.2 93.8% ±2.3 OpenHands Nov 2026 Official leaderboard
2 Gemini 4 Neon 93.7% ±2.3 OpenHands Nov 2026 Official leaderboard
3 Claude Sonnet 5.5 92.4% ±2.5 OpenHands Sep 2026 Official leaderboard
4 Gemini 4 Argon 91.9% ±3.7 OpenHands Sep 2026 Official leaderboard
5 Claude Haiku 5.5 90.4% ±3.4 OpenHands Oct 2026 Official leaderboard
6 Claude Fable 5 90.3% ±4.1 OpenHands Jun 2026 Official leaderboard
7 Claude Opus 5.5 90.3% ±3 OpenHands Sep 2026 Official leaderboard
8 Claude Fable 5.1 90.3% ±3.1 OpenHands Sep 2026 Official leaderboard
9 GPT-6.2 Astra 89.9% ±4 OpenHands Nov 2026 Official leaderboard
10 GPT-6 Astra 89.6% ±4.3 OpenHands Sep 2026 Official leaderboard
11 Gemini 4 Carbon 89.1% ±4.5 OpenHands Oct 2026 Official leaderboard
12 GPT-6.1 Sol 88.9% ±3.7 OpenHands Sep 2026 Official leaderboard
13 Claude Opus 5 88.4% ±5.9 OpenHands Jul 2026 Official leaderboard
14 GPT-6 Sol 87.8% ±5 OpenHands Sep 2026 Official leaderboard
15 Gemini 4 Hydrogen 87.8% ±4.5 OpenHands Nov 2026 Official leaderboard
16 DeepSeek V4.1 Flash 1017 87.7% ±5.4 OpenHands Oct 2026 Official leaderboard
17 GPT-6.2 Sol 87.6% ±4.3 OpenHands Nov 2026 Official leaderboard
18 Qwen4-Max 86.7% ±4.5 OpenHands Nov 2026 Official leaderboard
19 Grok 4.7 86.2% ±4.3 OpenHands Sep 2026 Official leaderboard
20 Muse Wonder Full 86% ±3.8 OpenHands Oct 2026 Official leaderboard
21 Muse Spark 1.3 85.9% ±4.9 max, OpenHands Sep 2026 Official leaderboard
22 MiMo-V2.6 Pro 85.2% ±6.6 OpenHands Sep 2026 Official leaderboard
23 Kimi K3 85% ±5.4 OpenHands Jul 2026 Official leaderboard
24 Kimi K3.1 Pro 84.9% ±5.8 OpenHands Oct 2026 Official leaderboard
25 DeepSeek V4.1 Flash 84.7% ±5.7 OpenHands Sep 2026 Official leaderboard
26 GPT-6.1 Terra 84% ±5.6 OpenHands Oct 2026 Official leaderboard
27 Ember-1 83.9% ±6.2 OpenHands Sep 2026 Official leaderboard
28 Muse Ember 83.7% ±5.4 OpenHands Nov 2026 Official leaderboard
29 Muse Wonder 83.1% ±4.9 OpenHands Oct 2026 Official leaderboard
30 Claude Opus 4.8 82.7% ±6 OpenHands May 2026 Official leaderboard
31 DeepSeek V4 Pro 0813 82.3% ±6.2 OpenHands Aug 2026 Official leaderboard
32 GPT-6 Luna 81.7% ±6.6 OpenHands Sep 2026 Official leaderboard
33 Mistral Large 4 81.6% ±6.4 OpenHands Oct 2026 Official leaderboard
34 Inkling 2 Small 81.5% ±6.8 OpenHands Nov 2026 Official leaderboard
35 Claude Sonnet 5 81.3% ±6 OpenHands Jun 2026 Official leaderboard
36 Tencent Hy4 81% ±7 OpenHands Nov 2026 Official leaderboard
37 GPT-5.6 Sol 80.5% ±7.3 OpenHands Jul 2026 Official leaderboard
38 Muse Spark 1.2 79.1% ±6.5 OpenHands Aug 2026 Official leaderboard
39 MiMo-V2.6 Flash 79% ±7.9 OpenHands Sep 2026 Official leaderboard
40 Gemini 3.8 Flash 78.7% ±7.6 OpenHands Sep 2026 Official leaderboard
41 Mistral Large 4 Preview 78.4% ±7 OpenHands Oct 2026 Official leaderboard
42 GLM-5.3 78.1% ±7.7 OpenHands Aug 2026 Official leaderboard
43 Tencent Hy4 preview 77.5% ±7.9 OpenHands Aug 2026 Official leaderboard
44 Qwen3.8 Max (0902) 77.1% ±7.3 OpenHands Sep 2026 Official leaderboard
45 GPT-5.6 Luna 77.1% ±6 OpenHands Jul 2026 Official leaderboard
46 Grok 4.6 76.2% ±7.5 OpenHands Aug 2026 Official leaderboard
47 Muse Spark Open 76.2% ±6.5 OpenHands Oct 2026 Official leaderboard
48 DeepSeek V4 Flash 0731 74.7% ±6.7 OpenHands Jul 2026 Official leaderboard
49 GPT-5.6 Terra 74.6% ±8.2 OpenHands Jul 2026 Official leaderboard
50 MiniMax-M3.1-Flash 73.9% ±8.3 OpenHands Oct 2026 Official leaderboard
51 Muse Spark 1.1 72.2% ±6.5 OpenHands Jul 2026 Official leaderboard
52 Claude Opus 4.7 71% ±8.8 OpenHands Apr 2026 Official leaderboard
53 Step 5 Preview 70.4% ±8.4 OpenHands Sep 2026 Official leaderboard
54 Gemini 3.7 Flash 70.4% ±9.5 OpenHands Aug 2026 Official leaderboard
55 GPT-5.5 69.8% ±8.9 OpenHands Apr 2026 Official leaderboard
56 Grok 4.5 69% ±8.8 OpenHands Jul 2026 Official leaderboard
57 GPT-5.4 67.4% ±9.5 xhigh, OpenHands Mar 2026 Official leaderboard
58 Beam 1.0 66.4% ±8 OpenHands Oct 2026 Official leaderboard
59 Qwen 3.8 27B 64.8% ±9.5 OpenHands Aug 2026 Official leaderboard
60 Qwen 3.8 Max 64.7% ±10.5 OpenHands Aug 2026 Official leaderboard
61 Gemini 3.6 Flash 64% ±8.4 OpenHands Jul 2026 Official leaderboard
62 GLM-5.2 64% ±9.4 OpenHands Jun 2026 Official leaderboard
63 GPT-5.3 Codex 61.8% ±9.2 OpenHands Feb 2026 Official leaderboard
64 Mercury 3 60.9% ±10.5 OpenHands Oct 2026 Official leaderboard
65 Claude Opus 4.6 57.6% ±8.6 non-thinking, OpenHands Feb 2026 Official leaderboard
66 Claude Sonnet 4.6 55.8% ±9.2 Claude Code Feb 2026 Official leaderboard
67 GPT-5.2 53.5% ±9.9 OpenHands Dec 2025 Official leaderboard
68 DeepSeek-V4-Pro 49.9% ±9.3 OpenHands Apr 2026 Official leaderboard
69 Composer 2.5 49.6% ±9.7 Cursor CLI May 2026 Official leaderboard
70 Gemini 3.5 Flash 48.7% ±9.3 OpenHands May 2026 Official leaderboard
71 GPT-5.4 Mini 48% ±11 OpenHands Mar 2026 Official leaderboard
72 Qwen3.7-Max 47.7% ±9.1 OpenHands May 2026 Official leaderboard
73 MiniMax-M3 47.6% ±10.7 OpenHands Jun 2026 Official leaderboard
74 Kimi K2.7 Code 47.2% ±10.2 OpenHands Jun 2026 Official leaderboard
75 Qwen3.7-Plus 46.4% ±9 OpenHands Jun 2026 Official leaderboard
76 MiMo-V2.5 42.2% ±9 OpenHands Apr 2026 Official leaderboard
77 GPT-5.2 Codex 37.9% ±9 OpenHands Dec 2025 Official leaderboard
78 Kimi K2.6 37.9% ±9.6 OpenHands Apr 2026 Official leaderboard
79 Gemini 3.5 Flash-Lite 37.2% ±9.1 OpenHands Jul 2026 Official leaderboard
80 MiMo-V2.5-Pro 34.1% ±8.9 OpenHands Apr 2026 Official leaderboard
81 Gemini 3.1 Pro 32% ±8.5 OpenHands Feb 2026 Official leaderboard
82 GLM-5.1 31.5% ±8.9 OpenHands Apr 2026 Official leaderboard
83 GLM-5.3-Flash 30.8% ±10.3 OpenHands Aug 2026 Official leaderboard
84 GPT-5.4 Nano 26.1% ±9.9 OpenHands Mar 2026 Official leaderboard
85 Qwen 3.6 Plus 25.6% ±7.7 OpenHands Mar 2026 Official leaderboard
86 GPT-5.1 24.6% ±8.3 OpenHands Nov 2025 Official leaderboard
87 GLM-5 23.4% ±7.9 thinking, OpenHands Feb 2026 Official leaderboard
88 Claude Sonnet 4.5 22.6% ±7.3 thinking, OpenHands Sep 2025 Official leaderboard
89 GPT-5.1-Codex-Max 22.2% ±7.5 OpenHands Nov 2025 Official leaderboard
90 Claude Opus 4.5 20.6% ±6.2 thinking, OpenHands Nov 2025 Official leaderboard
91 Gemini 3 Flash 20.2% ±7.7 OpenHands Dec 2025 Official leaderboard
92 GPT-5 20.1% ±6.7 OpenHands Aug 2025 Official leaderboard
93 Muse Spark 19.7% ±7.8 OpenHands Apr 2026 Official leaderboard
94 Grok 4.3 19.4% ±8.3 OpenHands Apr 2026 Official leaderboard
95 Inkling 19.2% ±7.2 OpenHands Jul 2026 Official leaderboard
96 Inkling-Small 19.1% ±6.7 OpenHands Jul 2026 Official leaderboard
97 Kimi K2.5 17.5% ±6.4 thinking, OpenHands Jan 2026 Official leaderboard
98 Qwen 3.5 Plus (hosted 397B-A17B) 15.7% ±6.2 thinking, OpenHands Feb 2026 Official leaderboard
99 MiniMax-M2.5 14.8% ±5.8 OpenHands Feb 2026 Official leaderboard
100 Gemini 3 Pro 14.3% ±6 OpenHands Nov 2025 Official leaderboard
101 GPT-5 mini 14.2% ±6.9 OpenHands Aug 2025 Official leaderboard
102 Grok Build 0.1 13.3% ±7.5 Grok Build May 2026 Official leaderboard
103 GPT-5.1-Codex 13.1% ±6.2 OpenHands Nov 2025 Official leaderboard
104 Qwen3.6 27B 11.9% ±6.7 OpenHands Apr 2026 Official leaderboard
105 MiniMax-M2.7 11.9% ±5.6 OpenHands Mar 2026 Official leaderboard
106 Claude Haiku 4.5 11.4% ±6.1 thinking, OpenHands Oct 2025 Official leaderboard
107 Laguna M.1 11% ±5.5 OpenHands Apr 2026 Official leaderboard
108 Nemotron 3 Ultra 7.64% ±5.5 OpenHands Jun 2026 Official leaderboard
109 Laguna XS.2 5.21% ±4.3 OpenHands Apr 2026 Official leaderboard
110 DeepSeek-V3.2 5.11% ±4.2 thinking, OpenHands Dec 2025 Official leaderboard
111 Grok 4.20 4.06% ±4 reasoning, OpenHands Feb 2026 Official leaderboard
112 Nemotron 3.5 Super VL 3.71% ±3.8 OpenHands Oct 2026 Official leaderboard
113 Mercury 2.5 3.68% ±2.8 OpenHands Sep 2026 Official leaderboard
114 Qwen3-Max 3.51% ±2.5 OpenHands Sep 2025 Official leaderboard
115 Ling-3.0-flash-Fin 3.40% ±2.5 OpenHands Sep 2026 Official leaderboard
116 GLM-4.6 3.09% ±2.3 OpenHands Sep 2025 Official leaderboard
117 Ling-3.0-flash 2.91% ±4.1 OpenHands Aug 2026 Official leaderboard
118 Mistral Medium 3.5 2.89% ±3.4 OpenHands Apr 2026 Official leaderboard
119 Grok 4.1 Fast 1.20% ±2.4 reasoning, OpenHands Nov 2025 Official leaderboard
120 Gemini 2.5 Pro Jun 2025 0.40% ±0.8 OpenHands Jun 2025 Official leaderboard
121 Grok 4 Fast 0% reasoning, OpenHands Sep 2025 Official leaderboard
122 Gemini 3.1 Flash-Lite 0% OpenHands Mar 2026 Official leaderboard
123 Mistral Small 4 0% OpenHands Mar 2026 Official leaderboard
124 Nemotron 3.5 Lightning 30B-A3B 0% OpenHands Aug 2026 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.2 93.8% OpenHands Nov 2026 Official leaderboard
2 Gemini 4 Neon 93.7% OpenHands Nov 2026 Official leaderboard
3 Claude Sonnet 5.5 92.4% OpenHands Sep 2026 Official leaderboard
4 Gemini 4 Argon 91.9% OpenHands Sep 2026 Official leaderboard
5 Claude Haiku 5.5 90.4% OpenHands Oct 2026 Official leaderboard
6 Claude Fable 5 90.3% OpenHands Jun 2026 Official leaderboard
7 Claude Opus 5.5 90.3% OpenHands Sep 2026 Official leaderboard
8 Claude Fable 5.1 90.3% OpenHands Sep 2026 Official leaderboard
9 GPT-6.2 Astra 89.9% OpenHands Nov 2026 Official leaderboard
10 GPT-6 Astra 89.6% OpenHands Sep 2026 Official leaderboard
11 Gemini 4 Carbon 89.1% OpenHands Oct 2026 Official leaderboard
12 GPT-6.1 Sol 88.9% OpenHands Sep 2026 Official leaderboard
13 Claude Opus 5 88.4% OpenHands Jul 2026 Official leaderboard
14 GPT-6 Sol 87.8% OpenHands Sep 2026 Official leaderboard
15 Gemini 4 Hydrogen 87.8% OpenHands Nov 2026 Official leaderboard
16 DeepSeek V4.1 Flash 1017 87.7% OpenHands Oct 2026 Official leaderboard
17 GPT-6.2 Sol 87.6% OpenHands Nov 2026 Official leaderboard
18 Qwen4-Max 86.7% OpenHands Nov 2026 Official leaderboard
19 Grok 4.7 86.2% OpenHands Sep 2026 Official leaderboard
20 Muse Wonder Full 86% OpenHands Oct 2026 Official leaderboard
21 Muse Spark 1.3 85.9% max, OpenHands Sep 2026 Official leaderboard
22 MiMo-V2.6 Pro 85.2% OpenHands Sep 2026 Official leaderboard
23 Kimi K3 85% OpenHands Jul 2026 Official leaderboard
24 Kimi K3.1 Pro 84.9% OpenHands Oct 2026 Official leaderboard
25 DeepSeek V4.1 Flash 84.7% OpenHands Sep 2026 Official leaderboard
26 GPT-6.1 Terra 84% OpenHands Oct 2026 Official leaderboard
27 Ember-1 83.9% OpenHands Sep 2026 Official leaderboard
28 Muse Ember 83.7% OpenHands Nov 2026 Official leaderboard
29 Muse Wonder 83.1% OpenHands Oct 2026 Official leaderboard
30 Claude Opus 4.8 82.7% OpenHands May 2026 Official leaderboard
31 DeepSeek V4 Pro 0813 82.3% OpenHands Aug 2026 Official leaderboard
32 GPT-6 Luna 81.7% OpenHands Sep 2026 Official leaderboard
33 Inkling 2 Small 81.6% Nov 2026 Lab model card
34 Mistral Large 4 81.6% OpenHands Oct 2026 Official leaderboard
35 Claude Sonnet 5 81.3% OpenHands Jun 2026 Official leaderboard
36 Tencent Hy4 81% OpenHands Nov 2026 Official leaderboard
37 GPT-5.6 Sol 80.5% OpenHands Jul 2026 Official leaderboard
38 Muse Spark 1.2 79.1% OpenHands Aug 2026 Official leaderboard
39 MiMo-V2.6 Flash 79% OpenHands Sep 2026 Official leaderboard
40 Gemini 3.8 Flash 78.7% OpenHands Sep 2026 Official leaderboard
41 Mistral Large 4 Preview 78.4% OpenHands Oct 2026 Official leaderboard
42 GLM-5.3 78.1% OpenHands Aug 2026 Official leaderboard
43 Inkling 2 Mini 77.9% Nov 2026 Lab model card
44 Tencent Hy4 preview 77.5% OpenHands Aug 2026 Official leaderboard
45 Qwen3.8 Max (0902) 77.1% OpenHands Sep 2026 Official leaderboard
46 GPT-5.6 Luna 77.1% OpenHands Jul 2026 Official leaderboard
47 Grok 4.6 76.2% OpenHands Aug 2026 Official leaderboard
48 Muse Spark Open 76.2% OpenHands Oct 2026 Official leaderboard
49 DeepSeek V4 Flash 0731 74.7% OpenHands Jul 2026 Official leaderboard
50 GPT-5.6 Terra 74.6% OpenHands Jul 2026 Official leaderboard
51 MiniMax-M3.1-Flash 73.9% OpenHands Oct 2026 Official leaderboard
52 Muse Spark 1.1 72.2% OpenHands Jul 2026 Official leaderboard
53 Claude Opus 4.7 71% OpenHands Apr 2026 Official leaderboard
54 Step 5 Preview 70.4% OpenHands Sep 2026 Official leaderboard
55 Gemini 3.7 Flash 70.4% OpenHands Aug 2026 Official leaderboard
56 GPT-5.5 69.8% OpenHands Apr 2026 Official leaderboard
57 Grok 4.5 69% OpenHands Jul 2026 Official leaderboard
58 GPT-5.4 67.4% xhigh, OpenHands Mar 2026 Official leaderboard
59 Beam 1.0 66.4% OpenHands Oct 2026 Official leaderboard
60 Qwen 3.8 27B 64.8% OpenHands Aug 2026 Official leaderboard
61 Qwen 3.8 Max 64.7% OpenHands Aug 2026 Official leaderboard
62 Gemini 3.6 Flash 64% OpenHands Jul 2026 Official leaderboard
63 GLM-5.2 64% OpenHands Jun 2026 Official leaderboard
64 GPT-5.3 Codex 61.8% OpenHands Feb 2026 Official leaderboard
65 Mercury 3 60.9% OpenHands Oct 2026 Official leaderboard
66 Claude Opus 4.6 57.6% non-thinking, OpenHands Feb 2026 Official leaderboard
67 Claude Sonnet 4.6 55.8% Claude Code Feb 2026 Official leaderboard
68 GPT-5.2 53.5% OpenHands Dec 2025 Official leaderboard
69 DeepSeek-V4-Pro 49.9% OpenHands Apr 2026 Official leaderboard
70 Composer 2.5 49.6% Cursor CLI May 2026 Official leaderboard
71 Gemini 3.5 Flash 48.7% OpenHands May 2026 Official leaderboard
72 GPT-5.4 Mini 48% OpenHands Mar 2026 Official leaderboard
73 Qwen3.7-Max 47.7% OpenHands May 2026 Official leaderboard
74 MiniMax-M3 47.6% OpenHands Jun 2026 Official leaderboard
75 Kimi K2.7 Code 47.2% OpenHands Jun 2026 Official leaderboard
76 Qwen3.7-Plus 46.4% OpenHands Jun 2026 Official leaderboard
77 MiMo-V2.5 42.2% OpenHands Apr 2026 Official leaderboard
78 GPT-5.2 Codex 37.9% OpenHands Dec 2025 Official leaderboard
79 Kimi K2.6 37.9% OpenHands Apr 2026 Official leaderboard
80 Gemini 3.5 Flash-Lite 37.2% OpenHands Jul 2026 Official leaderboard
81 MiMo-V2.5-Pro 34.1% OpenHands Apr 2026 Official leaderboard
82 Gemini 3.1 Pro 32% OpenHands Feb 2026 Official leaderboard
83 GLM-5.1 31.5% OpenHands Apr 2026 Official leaderboard
84 GLM-5.3-Flash 30.8% OpenHands Aug 2026 Official leaderboard
85 GPT-5.4 Nano 26.1% OpenHands Mar 2026 Official leaderboard
86 Qwen 3.6 Plus 25.6% OpenHands Mar 2026 Official leaderboard
87 GPT-5.1 24.6% OpenHands Nov 2025 Official leaderboard
88 GLM-5 23.4% thinking, OpenHands Feb 2026 Official leaderboard
89 Claude Sonnet 4.5 22.6% thinking, OpenHands Sep 2025 Official leaderboard
90 GPT-5.1-Codex-Max 22.2% OpenHands Nov 2025 Official leaderboard
91 Claude Opus 4.5 20.6% thinking, OpenHands Nov 2025 Official leaderboard
92 Gemini 3 Flash 20.2% OpenHands Dec 2025 Official leaderboard
93 GPT-5 20.1% OpenHands Aug 2025 Official leaderboard
94 Muse Spark 19.7% OpenHands Apr 2026 Official leaderboard
95 Grok 4.3 19.4% OpenHands Apr 2026 Official leaderboard
96 Inkling 19.2% OpenHands Jul 2026 Official leaderboard
97 Inkling-Small 19.1% OpenHands Jul 2026 Official leaderboard
98 Kimi K2.5 17.5% thinking, OpenHands Jan 2026 Official leaderboard
99 Qwen 3.5 Plus (hosted 397B-A17B) 15.7% thinking, OpenHands Feb 2026 Official leaderboard
100 MiniMax-M2.5 14.8% OpenHands Feb 2026 Official leaderboard
101 Gemini 3 Pro 14.3% OpenHands Nov 2025 Official leaderboard
102 GPT-5 mini 14.2% OpenHands Aug 2025 Official leaderboard
103 Grok Build 0.1 13.3% Grok Build May 2026 Official leaderboard
104 GPT-5.1-Codex 13.1% OpenHands Nov 2025 Official leaderboard
105 Qwen3.6 27B 11.9% OpenHands Apr 2026 Official leaderboard
106 MiniMax-M2.7 11.9% OpenHands Mar 2026 Official leaderboard
107 Claude Haiku 4.5 11.4% thinking, OpenHands Oct 2025 Official leaderboard
108 Laguna M.1 11% OpenHands Apr 2026 Official leaderboard
109 Nemotron 3 Ultra 7.64% OpenHands Jun 2026 Official leaderboard
110 Laguna XS.2 5.21% OpenHands Apr 2026 Official leaderboard
111 DeepSeek-V3.2 5.11% thinking, OpenHands Dec 2025 Official leaderboard
112 Grok 4.20 4.06% reasoning, OpenHands Feb 2026 Official leaderboard
113 Nemotron 3.5 Super VL 3.71% OpenHands Oct 2026 Official leaderboard
114 Mercury 2.5 3.68% OpenHands Sep 2026 Official leaderboard
115 Qwen3-Max 3.51% OpenHands Sep 2025 Official leaderboard
116 Ling-3.0-flash-Fin 3.40% OpenHands Sep 2026 Official leaderboard
117 GLM-4.6 3.09% OpenHands Sep 2025 Official leaderboard
118 Ling-3.0-flash 2.91% OpenHands Aug 2026 Official leaderboard
119 Mistral Medium 3.5 2.89% OpenHands Apr 2026 Official leaderboard
120 Grok 4.1 Fast 1.20% reasoning, OpenHands Nov 2025 Official leaderboard
121 Gemini 2.5 Pro Jun 2025 0.40% OpenHands Jun 2025 Official leaderboard
122 Grok 4 Fast 0% reasoning, OpenHands Sep 2025 Official leaderboard
123 Gemini 3.1 Flash-Lite 0% OpenHands Mar 2026 Official leaderboard
124 Mistral Small 4 0% OpenHands Mar 2026 Official leaderboard
125 Nemotron 3.5 Lightning 30B-A3B 0% OpenHands Aug 2026 Official leaderboard

Labs report scores for 3 models in their cards. Switch the scores above to rank them together.

Accuracy against latency

The Vals scatter chart with its x axis switched from Cost to Latency: mean time per test against accuracy, on a log scale. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

0% 20% 40% 60% 80% 100% 1001,00010K100K Latency (mean per test) (s), log scale, lower is better Accuracy Claude Fable 5.2Gemini 4 NeonGemini 4 HydrogenMuse Wonder FullMuse EmberBeam 1.0 0% 20% 40% 60% 80% 100% 1001,00010K100K Latency (mean per test) (s), log scale, lower is better Accuracy Claude Fable 5.2Gemini 4 NeonGemini 4 HydrogenMuse Wonder FullMuse EmberBeam 1.0
ModelSettingAccuracyLatency (mean per test)
Claude Fable 5.2 OpenHands 93.8% 2,437 s
Gemini 4 Neon OpenHands 93.7% 1,418 s
Claude Sonnet 5.5 OpenHands 92.4% 3,672 s
Gemini 4 Argon OpenHands 91.9% 2,248 s
Claude Haiku 5.5 OpenHands 90.4% 2,076 s
Claude Fable 5 OpenHands 90.4% 3,711 s
Claude Opus 5.5 OpenHands 90.3% 5,767 s
Claude Fable 5.1 OpenHands 90.3% 3,460 s
GPT-6.2 Astra OpenHands 89.9% 2,652 s
GPT-6 Astra OpenHands 89.6% 2,344 s
Gemini 4 Carbon OpenHands 89.1% 2,172 s
GPT-6.1 Sol OpenHands 88.9% 4,490 s
Claude Opus 5 OpenHands 88.4% 5,301 s
GPT-6 Sol OpenHands 87.8% 2,336 s
Gemini 4 Hydrogen OpenHands 87.8% 864.61 s
DeepSeek V4.1 Flash 1017 OpenHands 87.7% 973.83 s
GPT-6.2 Sol OpenHands 87.6% 2,871 s
Qwen4-Max OpenHands 86.7% 4,173 s
Grok 4.7 OpenHands 86.2% 2,147 s
Muse Wonder Full OpenHands 86% 648.34 s
Muse Spark 1.3 max, OpenHands 85.9% 977.67 s
MiMo-V2.6 Pro OpenHands 85.2% 3,647 s
Kimi K3 OpenHands 85% 999.02 s
Kimi K3.1 Pro OpenHands 84.9% 3,029 s
DeepSeek V4.1 Flash OpenHands 84.7% 928.1 s
GPT-6.1 Terra OpenHands 84% 2,250 s
Ember-1 OpenHands 83.9% 2,358 s
Muse Ember OpenHands 83.7% 167.86 s
Muse Wonder OpenHands 83.1% 350.08 s
Muse Spark 1.3 OpenHands 82.9% 773.8 s
Claude Opus 4.8 OpenHands 82.7% 4,549 s
DeepSeek V4 Pro 0813 OpenHands 82.3% 3,983 s
GPT-6 Luna OpenHands 81.7% 2,220 s
Mistral Large 4 OpenHands 81.6% 3,013 s
Inkling 2 Small OpenHands 81.6% 1,537 s
Claude Sonnet 5 OpenHands 81.3% 4,001 s
Tencent Hy4 OpenHands 81% 2,282 s
GPT-5.6 Sol OpenHands 80.5% 2,033 s
Muse Spark 1.2 OpenHands 79.1% 1,195 s
MiMo-V2.6 Flash OpenHands 79% 3,013 s
Gemini 3.8 Flash OpenHands 78.7% 519.1 s
Mistral Large 4 Preview OpenHands 78.4% 10,395 s
GLM-5.3 OpenHands 78.1% 3,847 s
Claude Opus 4.8 Claude Code 77.5% 1,142 s
Tencent Hy4 preview OpenHands 77.5% 2,493 s
Qwen3.8 Max (0902) OpenHands 77.1% 5,079 s
GPT-5.6 Luna OpenHands 77.1% 1,552 s
Grok 4.6 OpenHands 76.2% 1,530 s
Muse Spark Open OpenHands 76.2% 1,869 s
Hy4 Flash OpenHands 74.8% 1,008 s
DeepSeek V4 Flash 0731 OpenHands 74.7% 2,242 s
GPT-5.6 Terra OpenHands 74.6% 1,152 s
MiniMax-M3.1-Flash OpenHands 73.9% 1,264 s
Muse Spark 1.1 OpenHands 72.2% 615.43 s
Claude Opus 4.7 OpenHands 71% 2,142 s
Step 5 Preview OpenHands 70.4% 5,466 s
Gemini 3.7 Flash OpenHands 70.4% 1,424 s
GPT-5.5 OpenHands 69.9% 1,912 s
Grok 4.5 OpenHands 69% 830.06 s
GPT-5.4 xhigh, OpenHands 67.4% 5,175 s
GPT-5.5 Factory 67.4% 1,177 s
Beam 1.0 OpenHands 66.4% 145.26 s
Qwen 3.8 27B OpenHands 64.9% 2,619 s
Qwen 3.8 Max OpenHands 64.7% 10,135 s
Gemini 3.6 Flash OpenHands 64% 1,523 s
GLM-5.2 OpenHands 64% 3,748 s
GPT-5.3 Codex OpenHands 61.8% 4,548 s
Mercury 3 OpenHands 60.9% 566.58 s
GPT-5.5 Codex 58.2% 708.51 s
Claude Opus 4.6 non-thinking, OpenHands 57.6% 1,279 s
Claude Sonnet 4.6 Claude Code 55.8% 2,982 s
Claude Opus 4.6 thinking, OpenHands 53.5% 1,386 s
GPT-5.2 OpenHands 53.5% 4,971 s
Claude Sonnet 4.6 OpenHands 51.5% 1,572 s
DeepSeek-V4-Pro OpenHands 49.9% 3,419 s
Composer 2.5 Cursor CLI 49.6% 592.9 s
Gemini 3.5 Flash OpenHands 48.7% 893.07 s
GPT-5.4 Codex 48.5% 1,075 s
GPT-5.4 Mini OpenHands 48% 2,055 s
Qwen3.7-Max OpenHands 47.7% 4,690 s
MiniMax-M3 OpenHands 47.6% 5,073 s
Kimi K2.7 Code OpenHands 47.2% 10,318 s
Qwen3.7-Plus OpenHands 46.4% 2,232 s
MiMo-V2.5 OpenHands 42.2% 1,580 s
GPT-5.2 Codex OpenHands 37.9% 1,935 s
Kimi K2.6 OpenHands 37.9% 2,968 s
Gemini 3.5 Flash-Lite OpenHands 37.2% 501.52 s
MiMo-V2.5-Pro OpenHands 34.1% 2,404 s
Gemini 3.1 Pro OpenHands 32% 1,212 s
GLM-5.1 OpenHands 31.5% 2,015 s
GLM-5.3-Flash OpenHands 30.8% 5,027 s
GPT-5.4 Nano OpenHands 26.1% 3,276 s
Qwen 3.6 Plus OpenHands 25.6% 2,290 s
GPT-5.1 OpenHands 24.6% 1,836 s
GLM-5 thinking, OpenHands 23.4% 13,456 s
Claude Sonnet 4.5 thinking, OpenHands 22.6% 2,962 s
GPT-5.1-Codex-Max OpenHands 22.2% 31,183 s
Claude Opus 4.5 thinking, OpenHands 20.6% 2,283 s
Gemini 3 Flash OpenHands 20.2% 806.66 s
GPT-5 OpenHands 20.1% 1,852 s
Muse Spark OpenHands 19.7% 981.3 s
Grok 4.3 OpenHands 19.4% 589.41 s
Inkling OpenHands 19.2% 1,106 s
Inkling-Small OpenHands 19.1% 2,950 s
Kimi K2.5 thinking, OpenHands 17.5% 2,570 s
Qwen 3.5 Plus (hosted 397B-A17B) thinking, OpenHands 15.7% 3,016 s
MiniMax-M2.5 OpenHands 14.9% 3,066 s
Gemini 3 Pro OpenHands 14.3% 10,399 s
GPT-5 mini OpenHands 14.2% 698.25 s
Grok Build 0.1 Grok Build 13.4% 394.64 s
GPT-5.1-Codex OpenHands 13.1% 3,026 s
Qwen3.6 27B OpenHands 11.9% 9,763 s
MiniMax-M2.7 OpenHands 11.9% 1,377 s
Claude Haiku 4.5 thinking, OpenHands 11.4% 775.88 s
Laguna M.1 OpenHands 11% 2,382 s
Nemotron 3 Ultra OpenHands 7.6% 2,218 s
SWE-1.6 Fast Devin CLI 7.2% 189.13 s
Laguna XS.2 OpenHands 5.2% 2,199 s
DeepSeek-V3.2 thinking, OpenHands 5.1% 3,366 s
Grok 4.20 reasoning, OpenHands 4.1% 301.65 s
Nemotron 3.5 Super VL OpenHands 3.7% 262.9 s
Mercury 2.5 OpenHands 3.7% 193.04 s
Qwen3-Max OpenHands 3.5% 3,465 s
Ling-3.0-flash-Fin OpenHands 3.4% 1,451 s
GLM-4.6 OpenHands 3.1% 10,002 s
Ling-3.0-flash OpenHands 2.9% 1,006 s
Mistral Medium 3.5 OpenHands 2.9% 2,471 s
Grok 4.1 Fast reasoning, OpenHands 1.2% 527.56 s
Gemini 2.5 Pro Jun 2025 OpenHands 0.4% 2,097 s
Gemini 3.1 Flash-Lite OpenHands 0% 572.71 s
Nemotron 3.5 Lightning 30B-A3B OpenHands 0% 959.66 s
Grok 4 Fast reasoning, OpenHands 0% 1,449 s
Mistral Small 4 OpenHands 0% 1,815 s

Accuracy against latency

The Vals scatter chart with its x axis switched from Cost to Latency: mean time per test against accuracy, on a log scale. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

0% 20% 40% 60% 80% 100% 1001,00010K100K Latency (mean per test) (s), log scale, lower is better Accuracy Claude Fable 5.2Gemini 4 NeonGemini 4 HydrogenMuse Wonder FullMuse EmberBeam 1.0 0% 20% 40% 60% 80% 100% 1001,00010K100K Latency (mean per test) (s), log scale, lower is better Accuracy Claude Fable 5.2Gemini 4 NeonGemini 4 HydrogenMuse Wonder FullMuse EmberBeam 1.0
ModelSettingAccuracyLatency (mean per test)
Claude Fable 5.2 OpenHands 93.8% 2,437 s
Gemini 4 Neon OpenHands 93.7% 1,418 s
Claude Sonnet 5.5 OpenHands 92.4% 3,672 s
Gemini 4 Argon OpenHands 91.9% 2,248 s
Claude Haiku 5.5 OpenHands 90.4% 2,076 s
Claude Fable 5 OpenHands 90.4% 3,711 s
Claude Opus 5.5 OpenHands 90.3% 5,767 s
Claude Fable 5.1 OpenHands 90.3% 3,460 s
GPT-6.2 Astra OpenHands 89.9% 2,652 s
GPT-6 Astra OpenHands 89.6% 2,344 s
Gemini 4 Carbon OpenHands 89.1% 2,172 s
GPT-6.1 Sol OpenHands 88.9% 4,490 s
Claude Opus 5 OpenHands 88.4% 5,301 s
GPT-6 Sol OpenHands 87.8% 2,336 s
Gemini 4 Hydrogen OpenHands 87.8% 864.61 s
DeepSeek V4.1 Flash 1017 OpenHands 87.7% 973.83 s
GPT-6.2 Sol OpenHands 87.6% 2,871 s
Qwen4-Max OpenHands 86.7% 4,173 s
Grok 4.7 OpenHands 86.2% 2,147 s
Muse Wonder Full OpenHands 86% 648.34 s
Muse Spark 1.3 max, OpenHands 85.9% 977.67 s
MiMo-V2.6 Pro OpenHands 85.2% 3,647 s
Kimi K3 OpenHands 85% 999.02 s
Kimi K3.1 Pro OpenHands 84.9% 3,029 s
DeepSeek V4.1 Flash OpenHands 84.7% 928.1 s
GPT-6.1 Terra OpenHands 84% 2,250 s
Ember-1 OpenHands 83.9% 2,358 s
Muse Ember OpenHands 83.7% 167.86 s
Muse Wonder OpenHands 83.1% 350.08 s
Muse Spark 1.3 OpenHands 82.9% 773.8 s
Claude Opus 4.8 OpenHands 82.7% 4,549 s
DeepSeek V4 Pro 0813 OpenHands 82.3% 3,983 s
GPT-6 Luna OpenHands 81.7% 2,220 s
Mistral Large 4 OpenHands 81.6% 3,013 s
Inkling 2 Small OpenHands 81.6% 1,537 s
Claude Sonnet 5 OpenHands 81.3% 4,001 s
Tencent Hy4 OpenHands 81% 2,282 s
GPT-5.6 Sol OpenHands 80.5% 2,033 s
Muse Spark 1.2 OpenHands 79.1% 1,195 s
MiMo-V2.6 Flash OpenHands 79% 3,013 s
Gemini 3.8 Flash OpenHands 78.7% 519.1 s
Mistral Large 4 Preview OpenHands 78.4% 10,395 s
GLM-5.3 OpenHands 78.1% 3,847 s
Claude Opus 4.8 Claude Code 77.5% 1,142 s
Tencent Hy4 preview OpenHands 77.5% 2,493 s
Qwen3.8 Max (0902) OpenHands 77.1% 5,079 s
GPT-5.6 Luna OpenHands 77.1% 1,552 s
Grok 4.6 OpenHands 76.2% 1,530 s
Muse Spark Open OpenHands 76.2% 1,869 s
DeepSeek V4 Flash 0731 OpenHands 74.7% 2,242 s
GPT-5.6 Terra OpenHands 74.6% 1,152 s
MiniMax-M3.1-Flash OpenHands 73.9% 1,264 s
Muse Spark 1.1 OpenHands 72.2% 615.43 s
Claude Opus 4.7 OpenHands 71% 2,142 s
Step 5 Preview OpenHands 70.4% 5,466 s
Gemini 3.7 Flash OpenHands 70.4% 1,424 s
GPT-5.5 OpenHands 69.9% 1,912 s
Grok 4.5 OpenHands 69% 830.06 s
GPT-5.4 xhigh, OpenHands 67.4% 5,175 s
GPT-5.5 Factory 67.4% 1,177 s
Beam 1.0 OpenHands 66.4% 145.26 s
Qwen 3.8 27B OpenHands 64.9% 2,619 s
Qwen 3.8 Max OpenHands 64.7% 10,135 s
Gemini 3.6 Flash OpenHands 64% 1,523 s
GLM-5.2 OpenHands 64% 3,748 s
GPT-5.3 Codex OpenHands 61.8% 4,548 s
Mercury 3 OpenHands 60.9% 566.58 s
GPT-5.5 Codex 58.2% 708.51 s
Claude Opus 4.6 non-thinking, OpenHands 57.6% 1,279 s
Claude Sonnet 4.6 Claude Code 55.8% 2,982 s
Claude Opus 4.6 thinking, OpenHands 53.5% 1,386 s
GPT-5.2 OpenHands 53.5% 4,971 s
Claude Sonnet 4.6 OpenHands 51.5% 1,572 s
DeepSeek-V4-Pro OpenHands 49.9% 3,419 s
Composer 2.5 Cursor CLI 49.6% 592.9 s
Gemini 3.5 Flash OpenHands 48.7% 893.07 s
GPT-5.4 Codex 48.5% 1,075 s
GPT-5.4 Mini OpenHands 48% 2,055 s
Qwen3.7-Max OpenHands 47.7% 4,690 s
MiniMax-M3 OpenHands 47.6% 5,073 s
Kimi K2.7 Code OpenHands 47.2% 10,318 s
Qwen3.7-Plus OpenHands 46.4% 2,232 s
MiMo-V2.5 OpenHands 42.2% 1,580 s
GPT-5.2 Codex OpenHands 37.9% 1,935 s
Kimi K2.6 OpenHands 37.9% 2,968 s
Gemini 3.5 Flash-Lite OpenHands 37.2% 501.52 s
MiMo-V2.5-Pro OpenHands 34.1% 2,404 s
Gemini 3.1 Pro OpenHands 32% 1,212 s
GLM-5.1 OpenHands 31.5% 2,015 s
GLM-5.3-Flash OpenHands 30.8% 5,027 s
GPT-5.4 Nano OpenHands 26.1% 3,276 s
Qwen 3.6 Plus OpenHands 25.6% 2,290 s
GPT-5.1 OpenHands 24.6% 1,836 s
GLM-5 thinking, OpenHands 23.4% 13,456 s
Claude Sonnet 4.5 thinking, OpenHands 22.6% 2,962 s
GPT-5.1-Codex-Max OpenHands 22.2% 31,183 s
Claude Opus 4.5 thinking, OpenHands 20.6% 2,283 s
Gemini 3 Flash OpenHands 20.2% 806.66 s
GPT-5 OpenHands 20.1% 1,852 s
Muse Spark OpenHands 19.7% 981.3 s
Grok 4.3 OpenHands 19.4% 589.41 s
Inkling OpenHands 19.2% 1,106 s
Inkling-Small OpenHands 19.1% 2,950 s
Kimi K2.5 thinking, OpenHands 17.5% 2,570 s
Qwen 3.5 Plus (hosted 397B-A17B) thinking, OpenHands 15.7% 3,016 s
MiniMax-M2.5 OpenHands 14.9% 3,066 s
Gemini 3 Pro OpenHands 14.3% 10,399 s
GPT-5 mini OpenHands 14.2% 698.25 s
Grok Build 0.1 Grok Build 13.4% 394.64 s
GPT-5.1-Codex OpenHands 13.1% 3,026 s
Qwen3.6 27B OpenHands 11.9% 9,763 s
MiniMax-M2.7 OpenHands 11.9% 1,377 s
Claude Haiku 4.5 thinking, OpenHands 11.4% 775.88 s
Laguna M.1 OpenHands 11% 2,382 s
Nemotron 3 Ultra OpenHands 7.6% 2,218 s
Laguna XS.2 OpenHands 5.2% 2,199 s
DeepSeek-V3.2 thinking, OpenHands 5.1% 3,366 s
Grok 4.20 reasoning, OpenHands 4.1% 301.65 s
Nemotron 3.5 Super VL OpenHands 3.7% 262.9 s
Mercury 2.5 OpenHands 3.7% 193.04 s
Qwen3-Max OpenHands 3.5% 3,465 s
Ling-3.0-flash-Fin OpenHands 3.4% 1,451 s
GLM-4.6 OpenHands 3.1% 10,002 s
Ling-3.0-flash OpenHands 2.9% 1,006 s
Mistral Medium 3.5 OpenHands 2.9% 2,471 s
Grok 4.1 Fast reasoning, OpenHands 1.2% 527.56 s
Gemini 2.5 Pro Jun 2025 OpenHands 0.4% 2,097 s
Gemini 3.1 Flash-Lite OpenHands 0% 572.71 s
Nemotron 3.5 Lightning 30B-A3B OpenHands 0% 959.66 s
Grok 4 Fast reasoning, OpenHands 0% 1,449 s
Mistral Small 4 OpenHands 0% 1,815 s

The best score, 93.8%, costs $22 a task

Runs on the cost frontier score higher than any run that's cheaper. Claude Haiku 5.5 comes within 3.3 points of the best for $6.07 a task, 4 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

DeepSeek Anthropic Meta Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2 Gemini 4 Neon Gemini 4 Hydrogen DeepSeek V4.1 Flash 1017 Muse Ember 0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2 Gemini 4 Hydrogen

Source: Official Vibe Code Bench leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 OpenHands 93.8% $22 257,566
Gemini 4 Neon OpenHands 93.7% $6.76 159,652
Claude Sonnet 5.5 OpenHands 92.4% $20 509,821
Gemini 4 Argon OpenHands 91.9% $8.11 108,727
Claude Haiku 5.5 OpenHands 90.4% $6.07 477,279
Claude Fable 5 OpenHands 90.3% $42
Claude Opus 5.5 OpenHands 90.3% $36 394,450
Claude Fable 5.1 OpenHands 90.3% $33 224,587
GPT-6.2 Astra OpenHands 89.9% $7.92 93,921
GPT-6 Astra OpenHands 89.6% $39 153,442
Gemini 4 Carbon OpenHands 89.1% $8.25 144,698
GPT-6.1 Sol OpenHands 88.9% $6.23 203,736
Claude Opus 5 OpenHands 88.4% $34 195,337
GPT-6 Sol OpenHands 87.8% $26 136,469
Gemini 4 Hydrogen OpenHands 87.8% $4.55 184,074
DeepSeek V4.1 Flash 1017 OpenHands 87.7% $0.58 184,924
GPT-6.2 Sol OpenHands 87.6% $3.23 126,894
Qwen4-Max OpenHands 86.7% $41 114,443
Grok 4.7 OpenHands 86.2% $16 158,477
Muse Wonder Full OpenHands 86% $2.29 41,423
Muse Spark 1.3 max OpenHands 85.9% $2.54 60,971
MiMo-V2.6 Pro OpenHands 85.2% $1.04 149,562
Kimi K3 OpenHands 85% $10 100,216
Kimi K3.1 Pro OpenHands 84.9% $7.49 116,139
DeepSeek V4.1 Flash OpenHands 84.7% $0.41 143,729
GPT-6.1 Terra OpenHands 84% $6.00 118,574
Ember-1 OpenHands 83.9% $8.56 81,426
Muse Ember OpenHands 83.7% $0.05 37,758
Muse Wonder OpenHands 83.1% $1.80 36,607
Muse Spark 1.3 OpenHands 82.9% $2.10 54,278
Claude Opus 4.8 OpenHands 82.7% $27
DeepSeek V4 Pro 0813 OpenHands 82.3% $0.36 154,929
GPT-6 Luna OpenHands 81.7% $1.35 180,091
Mistral Large 4 OpenHands 81.6% $12 127,093
Inkling 2 Small OpenHands 81.5% $1.90 188,867
Claude Sonnet 5 OpenHands 81.3% $25
Tencent Hy4 OpenHands 81% $2.39 154,239
GPT-5.6 Sol OpenHands 80.5% $33
Muse Spark 1.2 OpenHands 79.1% $1.53 47,024
MiMo-V2.6 Flash OpenHands 79% $0.56 165,570
Gemini 3.8 Flash OpenHands 78.7% $6.87 122,884
Mistral Large 4 Preview OpenHands 78.4% $20 161,302
GLM-5.3 OpenHands 78.1% $12 163,467
Claude Opus 4.8 Claude Code 77.5% $6.86 80,284
Tencent Hy4 preview OpenHands 77.5% $2.18 125,264
Qwen3.8 Max (0902) OpenHands 77.1% $12 166,554
GPT-5.6 Luna OpenHands 77.1% $0.73 73,204
Grok 4.6 OpenHands 76.2% $4.88 76,097
Muse Spark Open OpenHands 76.2% $1.66 41,718
DeepSeek V4 Flash 0731 OpenHands 74.7% $0.91 155,038
GPT-5.6 Terra OpenHands 74.6% $7.89 97,307
MiniMax-M3.1-Flash OpenHands 73.9% $1.15 169,763
Muse Spark 1.1 OpenHands 72.2% $0.92
Claude Opus 4.7 OpenHands 71% $21
Step 5 Preview OpenHands 70.4% $5.86 163,654
Gemini 3.7 Flash OpenHands 70.4% $4.83 117,608
GPT-5.5 OpenHands 69.8% $17 61,798
Grok 4.5 OpenHands 69% $4.10
GPT-5.4 xhigh OpenHands 67.4% $11 116,452
GPT-5.5 Factory 67.4% $5.93 48,868
Beam 1.0 OpenHands 66.4% 40,097
Qwen 3.8 27B OpenHands 64.8% $23 176,498
Qwen 3.8 Max OpenHands 64.7% $8.24 119,118
Gemini 3.6 Flash OpenHands 64% $3.04 81,300
GLM-5.2 OpenHands 64% $8.49
GPT-5.3 Codex OpenHands 61.8% $12 80,188
Mercury 3 OpenHands 60.9% $0.63 129,860
GPT-5.5 Codex 58.2% $3.63 40,368
Claude Opus 4.6 non-thinking OpenHands 57.6% $8.69
Claude Sonnet 4.6 Claude Code 55.8% $7.35 7,242
GPT-5.2 OpenHands 53.5% $18 88,133
Claude Opus 4.6 thinking OpenHands 53.5% $8.28
Claude Sonnet 4.6 OpenHands 51.5% $5.91
DeepSeek-V4-Pro OpenHands 49.9% $2.21
Composer 2.5 Cursor CLI 49.6% $0.50 40,586
Gemini 3.5 Flash OpenHands 48.7% $2.54 65,703
GPT-5.4 Codex 48.5% $7.51 72,205
GPT-5.4 Mini OpenHands 48% $1.19
Qwen3.7-Max OpenHands 47.7% $11 47,095
MiniMax-M3 OpenHands 47.6% $6.45
Kimi K2.7 Code OpenHands 47.2% $3.84
Qwen3.7-Plus OpenHands 46.4% $1.08
MiMo-V2.5 OpenHands 42.2% $0.07
GPT-5.2 Codex OpenHands 37.9% $4.15 53,174
Kimi K2.6 OpenHands 37.9% $1.93
Gemini 3.5 Flash-Lite OpenHands 37.2% $0.83 67,200
MiMo-V2.5-Pro OpenHands 34.1% $0.17
Gemini 3.1 Pro OpenHands 32% $3.83
GLM-5.1 OpenHands 31.5% $2.89
GLM-5.3-Flash OpenHands 30.8% $4.70 238,331
GPT-5.4 Nano OpenHands 26.1% $1.28
Qwen 3.6 Plus OpenHands 25.6% $5.45 53,075
GPT-5.1 OpenHands 24.6% $2.57
GLM-5 thinking OpenHands 23.4% $40 57,419
Claude Sonnet 4.5 thinking OpenHands 22.6% $6.66
GPT-5.1-Codex-Max OpenHands 22.2% $7.32
Claude Opus 4.5 thinking OpenHands 20.6% $33
Gemini 3 Flash OpenHands 20.2% $0.94
GPT-5 OpenHands 20.1% $1.53
Muse Spark OpenHands 19.7% 41,618
Grok 4.3 OpenHands 19.4% $1.28
Inkling OpenHands 19.2% $1.61 45,766
Inkling-Small OpenHands 19.1% $0.43 42,713
Kimi K2.5 thinking OpenHands 17.5% $0.88
Qwen 3.5 Plus (hosted 397B-A17B) thinking OpenHands 15.7% $3.80 48,819
MiniMax-M2.5 OpenHands 14.8% $2.20 52,563
Gemini 3 Pro OpenHands 14.3% $6.42
GPT-5 mini OpenHands 14.2% $0.25 21,744
Grok Build 0.1 Grok Build 13.3%
GPT-5.1-Codex OpenHands 13.1% $3.80
Qwen3.6 27B OpenHands 11.9% $4.66 51,528
MiniMax-M2.7 OpenHands 11.9% $2.82
Claude Haiku 4.5 thinking OpenHands 11.4% $1.31
Laguna M.1 OpenHands 11% 62,251
Nemotron 3 Ultra OpenHands 7.64% 108,817
Laguna XS.2 OpenHands 5.21% 51,638
DeepSeek-V3.2 thinking OpenHands 5.11% $2.47
Grok 4.20 reasoning OpenHands 4.06% $0.77
Nemotron 3.5 Super VL OpenHands 3.71% $0.54 58,038
Mercury 2.5 OpenHands 3.68% $0.38 28,520
Qwen3-Max OpenHands 3.51% $6.02
Ling-3.0-flash-Fin OpenHands 3.40% $0.23 147,021
GLM-4.6 OpenHands 3.09% $11
Ling-3.0-flash OpenHands 2.91% $0.29 76,123
Mistral Medium 3.5 OpenHands 2.89% $17 78,622
Grok 4.1 Fast reasoning OpenHands 1.20% $0.21 16,637
Gemini 2.5 Pro Jun 2025 OpenHands 0.40% $1.22
Grok 4 Fast reasoning OpenHands 0% $0.13
Gemini 3.1 Flash-Lite OpenHands 0% $0.51 20,850
Mistral Small 4 OpenHands 0% $6.47 60,285
Nemotron 3.5 Lightning 30B-A3B OpenHands 0% $0.13 66,932

The best score, 93.8%, costs $22 a task

Runs on the cost frontier score higher than any run that's cheaper. Claude Haiku 5.5 comes within 3.3 points of the best for $6.07 a task, 4 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

DeepSeek Anthropic Meta Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2 Gemini 4 Neon Gemini 4 Hydrogen DeepSeek V4.1 Flash 1017 Muse Ember 0% 20% 40% 60% 80% 100% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2 Gemini 4 Hydrogen

Source: Official Vibe Code Bench leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 OpenHands 93.8% $22 257,566
Gemini 4 Neon OpenHands 93.7% $6.76 159,652
Claude Sonnet 5.5 OpenHands 92.4% $20 509,821
Gemini 4 Argon OpenHands 91.9% $8.11 108,727
Claude Haiku 5.5 OpenHands 90.4% $6.07 477,279
Claude Fable 5 OpenHands 90.3% $42
Claude Opus 5.5 OpenHands 90.3% $36 394,450
Claude Fable 5.1 OpenHands 90.3% $33 224,587
GPT-6.2 Astra OpenHands 89.9% $7.92 93,921
GPT-6 Astra OpenHands 89.6% $39 153,442
Gemini 4 Carbon OpenHands 89.1% $8.25 144,698
GPT-6.1 Sol OpenHands 88.9% $6.23 203,736
Claude Opus 5 OpenHands 88.4% $34 195,337
GPT-6 Sol OpenHands 87.8% $26 136,469
Gemini 4 Hydrogen OpenHands 87.8% $4.55 184,074
DeepSeek V4.1 Flash 1017 OpenHands 87.7% $0.58 184,924
GPT-6.2 Sol OpenHands 87.6% $3.23 126,894
Qwen4-Max OpenHands 86.7% $41 114,443
Grok 4.7 OpenHands 86.2% $16 158,477
Muse Wonder Full OpenHands 86% $2.29 41,423
Muse Spark 1.3 max OpenHands 85.9% $2.54 60,971
MiMo-V2.6 Pro OpenHands 85.2% $1.04 149,562
Kimi K3 OpenHands 85% $10 100,216
Kimi K3.1 Pro OpenHands 84.9% $7.49 116,139
DeepSeek V4.1 Flash OpenHands 84.7% $0.41 143,729
GPT-6.1 Terra OpenHands 84% $6.00 118,574
Ember-1 OpenHands 83.9% $8.56 81,426
Muse Ember OpenHands 83.7% $0.05 37,758
Muse Wonder OpenHands 83.1% $1.80 36,607
Muse Spark 1.3 OpenHands 82.9% $2.10 54,278
Claude Opus 4.8 OpenHands 82.7% $27
DeepSeek V4 Pro 0813 OpenHands 82.3% $0.36 154,929
GPT-6 Luna OpenHands 81.7% $1.35 180,091
Mistral Large 4 OpenHands 81.6% $12 127,093
Inkling 2 Small OpenHands 81.5% $1.90 188,867
Claude Sonnet 5 OpenHands 81.3% $25
Tencent Hy4 OpenHands 81% $2.39 154,239
GPT-5.6 Sol OpenHands 80.5% $33
Muse Spark 1.2 OpenHands 79.1% $1.53 47,024
MiMo-V2.6 Flash OpenHands 79% $0.56 165,570
Gemini 3.8 Flash OpenHands 78.7% $6.87 122,884
Mistral Large 4 Preview OpenHands 78.4% $20 161,302
GLM-5.3 OpenHands 78.1% $12 163,467
Claude Opus 4.8 Claude Code 77.5% $6.86 80,284
Tencent Hy4 preview OpenHands 77.5% $2.18 125,264
Qwen3.8 Max (0902) OpenHands 77.1% $12 166,554
GPT-5.6 Luna OpenHands 77.1% $0.73 73,204
Grok 4.6 OpenHands 76.2% $4.88 76,097
Muse Spark Open OpenHands 76.2% $1.66 41,718
DeepSeek V4 Flash 0731 OpenHands 74.7% $0.91 155,038
GPT-5.6 Terra OpenHands 74.6% $7.89 97,307
MiniMax-M3.1-Flash OpenHands 73.9% $1.15 169,763
Muse Spark 1.1 OpenHands 72.2% $0.92
Claude Opus 4.7 OpenHands 71% $21
Step 5 Preview OpenHands 70.4% $5.86 163,654
Gemini 3.7 Flash OpenHands 70.4% $4.83 117,608
GPT-5.5 OpenHands 69.8% $17 61,798
Grok 4.5 OpenHands 69% $4.10
GPT-5.4 xhigh OpenHands 67.4% $11 116,452
GPT-5.5 Factory 67.4% $5.93 48,868
Beam 1.0 OpenHands 66.4% 40,097
Qwen 3.8 27B OpenHands 64.8% $23 176,498
Qwen 3.8 Max OpenHands 64.7% $8.24 119,118
Gemini 3.6 Flash OpenHands 64% $3.04 81,300
GLM-5.2 OpenHands 64% $8.49
GPT-5.3 Codex OpenHands 61.8% $12 80,188
Mercury 3 OpenHands 60.9% $0.63 129,860
GPT-5.5 Codex 58.2% $3.63 40,368
Claude Opus 4.6 non-thinking OpenHands 57.6% $8.69
Claude Sonnet 4.6 Claude Code 55.8% $7.35 7,242
GPT-5.2 OpenHands 53.5% $18 88,133
Claude Opus 4.6 thinking OpenHands 53.5% $8.28
Claude Sonnet 4.6 OpenHands 51.5% $5.91
DeepSeek-V4-Pro OpenHands 49.9% $2.21
Composer 2.5 Cursor CLI 49.6% $0.50 40,586
Gemini 3.5 Flash OpenHands 48.7% $2.54 65,703
GPT-5.4 Codex 48.5% $7.51 72,205
GPT-5.4 Mini OpenHands 48% $1.19
Qwen3.7-Max OpenHands 47.7% $11 47,095
MiniMax-M3 OpenHands 47.6% $6.45
Kimi K2.7 Code OpenHands 47.2% $3.84
Qwen3.7-Plus OpenHands 46.4% $1.08
MiMo-V2.5 OpenHands 42.2% $0.07
GPT-5.2 Codex OpenHands 37.9% $4.15 53,174
Kimi K2.6 OpenHands 37.9% $1.93
Gemini 3.5 Flash-Lite OpenHands 37.2% $0.83 67,200
MiMo-V2.5-Pro OpenHands 34.1% $0.17
Gemini 3.1 Pro OpenHands 32% $3.83
GLM-5.1 OpenHands 31.5% $2.89
GLM-5.3-Flash OpenHands 30.8% $4.70 238,331
GPT-5.4 Nano OpenHands 26.1% $1.28
Qwen 3.6 Plus OpenHands 25.6% $5.45 53,075
GPT-5.1 OpenHands 24.6% $2.57
GLM-5 thinking OpenHands 23.4% $40 57,419
Claude Sonnet 4.5 thinking OpenHands 22.6% $6.66
GPT-5.1-Codex-Max OpenHands 22.2% $7.32
Claude Opus 4.5 thinking OpenHands 20.6% $33
Gemini 3 Flash OpenHands 20.2% $0.94
GPT-5 OpenHands 20.1% $1.53
Muse Spark OpenHands 19.7% 41,618
Grok 4.3 OpenHands 19.4% $1.28
Inkling OpenHands 19.2% $1.61 45,766
Inkling-Small OpenHands 19.1% $0.43 42,713
Kimi K2.5 thinking OpenHands 17.5% $0.88
Qwen 3.5 Plus (hosted 397B-A17B) thinking OpenHands 15.7% $3.80 48,819
MiniMax-M2.5 OpenHands 14.8% $2.20 52,563
Gemini 3 Pro OpenHands 14.3% $6.42
GPT-5 mini OpenHands 14.2% $0.25 21,744
Grok Build 0.1 Grok Build 13.3%
GPT-5.1-Codex OpenHands 13.1% $3.80
Qwen3.6 27B OpenHands 11.9% $4.66 51,528
MiniMax-M2.7 OpenHands 11.9% $2.82
Claude Haiku 4.5 thinking OpenHands 11.4% $1.31
Laguna M.1 OpenHands 11% 62,251
Nemotron 3 Ultra OpenHands 7.64% 108,817
Laguna XS.2 OpenHands 5.21% 51,638
DeepSeek-V3.2 thinking OpenHands 5.11% $2.47
Grok 4.20 reasoning OpenHands 4.06% $0.77
Nemotron 3.5 Super VL OpenHands 3.71% $0.54 58,038
Mercury 2.5 OpenHands 3.68% $0.38 28,520
Qwen3-Max OpenHands 3.51% $6.02
Ling-3.0-flash-Fin OpenHands 3.40% $0.23 147,021
GLM-4.6 OpenHands 3.09% $11
Ling-3.0-flash OpenHands 2.91% $0.29 76,123
Mistral Medium 3.5 OpenHands 2.89% $17 78,622
Grok 4.1 Fast reasoning OpenHands 1.20% $0.21 16,637
Gemini 2.5 Pro Jun 2025 OpenHands 0.40% $1.22
Grok 4 Fast reasoning OpenHands 0% $0.13
Gemini 3.1 Flash-Lite OpenHands 0% $0.51 20,850
Mistral Small 4 OpenHands 0% $6.47 60,285
Nemotron 3.5 Lightning 30B-A3B OpenHands 0% $0.13 66,932

7 models have held number one on the official leaderboard since its release in March 2026 7 models have held number one with claims included since its release in March 2026

Lab claims count from the day the card came out, independent runs from the model’s release. We draw one row each for the benchmark’s own leaderboard and what labs claim in their cards.

OpenAI Anthropic Google DeepMind
Own leaderboard Passed 85% in Jun 2026, by Claude Fable 5
Lab claims Passed 85% in Oct 2026, by Gemini 4 Argon

Sources: Official Vibe Code Bench leaderboard and labs’ model cards

Claude Fable 5 on the official leaderboard passed 85% in June 2026

That was 3 months after its release. By the same age, 23% of the benchmarks from 2026 had passed 85%. 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 Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Fable 5 Claude Opus 4.8 Claude Opus 4.7 GPT-5.3 Codex GPT-5.2 GPT-5.1 0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Opus 4.8

Sources: Official Vibe Code Bench leaderboard and labs’ model cards

2 labs put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Google DeepMind
Thinking Machines
DateLaunch postScoreAs the post writes it
Nov 20, 2026 Inkling 2 Mini 77.9 Vibe Code Bench v1.1 (openhands, Vals)
Nov 5, 2026 Inkling 2 Small 81.6 Vibe Code Bench v1.1
Oct 13, 2026 Gemini 4 Argon 91.9 Vibe Code Bench

2 system cards report it, from 2 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Nov 20, 2026 Inkling 2 Mini model card 77.9 Vibe Code Bench v1.1 (openhands, Vals) Capabilities
Oct 13, 2026 Gemini 4 Argon model card 91.9 Vibe Code Bench (results table, p.5) Capabilities

2 lab claims next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 2 of them match, which usually means one side copied the other. Every benchmark.

Model Lab says Board says Gap
Inkling 2 Small 81.6% Vibe Code Bench v1.1 81.5% Leaderboard, OpenHands Same
Gemini 4 Argon 91.9% Vibe Code Bench 91.9% Leaderboard, OpenHands Same

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

Anthropic OpenAI
Vibe Code Bench Passed 85% in Jun 2026, by Claude Fable 5
Vibe Code Bench v1 Replaced by Bench, Mar 2026
GPT-5.2 41.3%
Vibe Code Bench Passed 85% in Jun 2026, by Claude Fable 5

Source: Official leaderboards

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

10 of 11 checks pass, 1 warning. Look first at the Confidence check. Likely. Some facts rest on secondary sources. Every benchmark

Catalog record 1 warning 1 pass
  • Pass Task access The tasks are private.
  • Warning Confidence Likely. Some facts rest on secondary sources.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (126 entries), launch posts (3 scores) and system cards (2 scores).
  • Pass Models scored 125 models we track, plus 2 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Inkling 2 Mini.
  • Pass Frontier models 10 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Note Headroom Saturated. The best is 93.8%, past 85% since June 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on vals.ai: 131 rows for models we track and 2 other entries at the last read.
  • Pass Last read November 19th, 2026.
  • Pass Read worked The last read worked.
Consistency 2 pass
  • Pass Lab claims Both pairs agree within 10 points.
  • Pass Known issues No published audit found problems.

How we check