Benchmark Almanac

Longform Creative Writing (EQ-Bench)

Claude Opus 5 86.3 leads since Jul 2026 Not in model cards

Models plan, revise and then write a short story or novella over eight 1,000-word turns from a minimal prompt. A judge model scores quality, and slop, repetition and degradation are tracked.

Not the in-house Moonshot Longform Writing eval with the same short name. v1.1 (2025-08-08) moved the judge to Claude Sonnet 4 and v1.11 (2026-02-19) to Sonnet 4.6.

Made by EQ-Bench (Sam Paech) Released May 2025 Creative writing All tasks public Last model added Oct 2026 Scores refreshed 2 Nov 2026
Level
University Undergraduates and entry-level professionals
Scored
Graded by a model A language model scores each response, usually against a rubric.
Answers
Open-ended Free-form text, documents, designs or media judged as a whole.
EQ-bench/longform-writing-bench 47 stars Leaderboard eqbench.com
140
Entries on its own leaderboard, 119 models we track

Claude Opus 5 leads with 86.3

Best score per entry from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

140 entries. Scroll sideways for the rest.

  1. 86.3 Claude Opus 5
  2. 85.3 Claude Fable 5.1
  3. 84.6 Claude Opus 5.5
  4. 83.0 Claude Fable 5
  5. 83.0 Grok 4.7
  6. 82.8 Muse Spark 1.3
  7. 82.8 GPT-6 Astra
  8. 82.8 GPT-6 Sol
  9. 82.1 aion-3.5
  10. 81.8 Claude Opus 4.7
  11. 81.8 GLM-5.3
  12. 81.7 GPT-5.6 Sol
  13. 81.5 Muse Spark 1.2
  14. 80.9 Gemini 4 Argon
  15. 80.8 Claude Opus 4.8
  16. 79.9 ox-alpha
  17. 79.9 Claude Sonnet 4.6
  18. 79.9 Kimi K3.1 Pro
  19. 79.6 Kimi K3
  20. 78.5 Kimi K2.6
  21. 78.3 GPT-5.4
  22. 78.3 Claude Sonnet 5
  23. 78.2 GPT-5.5
  24. 78.0 GPT-5.6 Terra
  25. 77.9 GLM-5.2
  26. 77.7 Claude Opus 4.6
  27. 76.8 Gemini 3.8 Flash
  28. 75.6 DeepSeek-V4-Pro
  29. 75.5 GPT-5.6 Luna
  30. 74.9 Kimi K2.5
  31. 74.2 Altworld/Hemmingway-1
  32. 74.0 DeepSeek V4.1 Flash
  33. 73.7 MiMo-V2.5-Pro
  34. 73.5 GLM-5.1
  35. 73.1 gpt-5-2025-08-07-high-reasoning-high-reasoning
  36. 73.1 Claude Opus 4.5
  37. 73.0 GPT-5.4 Mini
  38. 72.8 GPT-5
  39. 72.5 Inkling
  40. 71.8 Gemini 3.5 Flash
  41. 71.6 GPT-5.2
  42. 71.4 Claude Sonnet 4.5
  43. 71.2 horizon-alpha
  44. 70.9 GLM-5
  45. 70.9 MiMo-V2-Pro
  46. 70.6 Muse Spark 1.1
  47. 70.5 Claude Opus 4.1
  48. 69.0 Kimi K2 Thinking
  49. 68.8 Gemini 3 Pro
  50. 68.5 Grok 4.20
  51. 68.3 Qwen3.5 397B-A17B
  52. 68.2 Gemini 3.1 Pro
  53. 68.1 Kimi K2
  54. 66.8 GPT-5 Chat
  55. 66.8 DeepSeek-V3.2
  56. 66.2 Qwen3-Max
  57. 66.0 GLM-4.7
  58. 66.0 DeepSeek-V4-Flash
  59. 65.3 MiniMax-M2.5
  60. 65.0 Claude Haiku 4.5
  61. 63.8 Kimi K2
  62. 62.7 hunter-alpha
  63. 62.7 o3
  64. 62.6 GPT-5.3 Instant
  65. 62.3 DeepSeek-V3.1
  66. 61.2 DeepSeek V4 Flash 0731
  67. 60.2 GPT-4o
  68. 60.2 DeepSeek-V3.2-Exp
  69. 60.1 DeepSeek-V3
  70. 59.5 DeepSeek-R1
  71. 59.4 tngtech/DeepSeek-R1T-Chimera
  72. 59.0 Qwen3.5-27B
  73. 58.8 Gemini 2.5 Pro
  74. 58.4 Claude 3.5 Sonnet
  75. 58.4 Claude Sonnet 4
  76. 57.8 MiniMax-M2.1
  77. 57.8 GPT-5.4 Nano
  78. 57.7 Claude 3.7 Sonnet
  79. 57.4 Qwen3-235B-A22B-Instruct
  80. 57.3 GLM-4.6
  81. 56.5 Gemma 4 31B IT
  82. 56.2 inclusionai/ling-1t
  83. 56.2 GPT-5 mini
  84. 55.7 Gemini 3 Flash
  85. 55.5 GLM-4.5
  86. 54.3 Qwen3.5-122B-A10B
  87. 53.3 Qwen 3.8 27B
  88. 53.1 MiMo-V2-Flash
  89. 52.4 AlexWortega/Qwen3.8-27B-antislop
  90. 51.2 Hermes 4 405B
  91. 51.0 healer-alpha
  92. 50.7 Gemma 4 26B A4B
  93. 49.0 Gemini 3.1 Flash-Lite
  94. 48.8 DeepSeek-V3.2-Speciale
  95. 48.7 Gemma 4 12B
  96. 48.6 Gemini 2.5 Flash
  97. 48.3 Mistral Large 3
  98. 47.8 glm-4.7-flash
  99. 47.3 bytedance-seed-1.6
  100. 47.2 openrouter/sherlock-dash-alpha
  101. 47.0 Olmo 3.1 32B Think
  102. 46.4 openrouter/bert-nebulon-alpha
  103. 46.4 MiniMax-M2
  104. 46.2 Mistral Medium 3.1
  105. 45.7 Grok 4
  106. 45.2 sam-paech/gemma-3-27b-it-antislop
  107. 44.5 Qwen3.5-35B-A3B
  108. 43.2 Mistral Small 4
  109. 42.9 sonoma-sky-alpha
  110. 41.9 Gemma 3 27B
  111. 41.8 Gemma 3 12B
  112. 41.6 Mistral Small 3.2
  113. 41.2 Reka Flash 3
  114. 40.5 Qwen3-32B
  115. 40.1 Grok 4.1 Fast
  116. 39.7 ftpo-exp501
  117. 38.8 Nova 2 Lite
  118. 36.9 QwQ-32B
  119. 35.9 Qwen3-14B
  120. 35.8 gpt-oss-120b
  121. 35.6 Qwen3-30B-A3B
  122. 35.1 LongCat-Flash-Chat
  123. 35.0 Trinity Large Preview
  124. 34.8 minimax-m2-her
  125. 34.4 Gemma 3 4B
  126. 32.9 nvidia/nemotron-3-nano-30b-a3b:free
  127. 32.6 mistral-small-creative
  128. 32.4 GPT-4.1 nano
  129. 31.9 Llama 4 Maverick
  130. 31.9 Ministral 3 14B
  131. 31.5 Nanbeige/Nanbeige4-3B-Thinking-2511
  132. 31.5 Qwen3-Next-80B-A3B-Instruct
  133. 31.1 Llama 4 Scout
  134. 29.4 openrouter/aurora-alpha
  135. 28.7 Mistral NeMo
  136. 28.0 GPT-5 nano
  137. 27.0 Solar Pro 3
  138. 22.8 Phi-4-multimodal
  139. 21.2 gpt-oss-20b
  140. 17.0 Qwen3-4B
  1. 1 Claude Opus 5 86.3
  2. 2 Claude Fable 5.1 85.3
  3. 3 Claude Opus 5.5 84.6
  4. 4 Claude Fable 5 83.0
  5. 5 Grok 4.7 83.0
  6. 6 Muse Spark 1.3 82.8
  7. 7 GPT-6 Astra 82.8
  8. 8 GPT-6 Sol 82.8
  9. 9 aion-3.5 82.1
  10. 10 Claude Opus 4.7 81.8
  11. 11 GLM-5.3 81.8
  12. 12 GPT-5.6 Sol 81.7
  13. 13 Muse Spark 1.2 81.5
  14. 14 Gemini 4 Argon 80.9
  15. 15 Claude Opus 4.8 80.8
  16. 16 ox-alpha 79.9
  17. 17 Claude Sonnet 4.6 79.9
  18. 18 Kimi K3.1 Pro 79.9
  19. 19 Kimi K3 79.6
  20. 20 Kimi K2.6 78.5
  21. 21 GPT-5.4 78.3
  22. 22 Claude Sonnet 5 78.3
  23. 23 GPT-5.5 78.2
  24. 24 GPT-5.6 Terra 78.0
  25. 25 GLM-5.2 77.9
  26. 26 Claude Opus 4.6 77.7
  27. 27 Gemini 3.8 Flash 76.8
  28. 28 DeepSeek-V4-Pro 75.6
  29. 29 GPT-5.6 Luna 75.5
  30. 30 Kimi K2.5 74.9
  31. 31 Altworld/Hemmingway-1 74.2
  32. 32 DeepSeek V4.1 Flash 74.0
  33. 33 MiMo-V2.5-Pro 73.7
  34. 34 GLM-5.1 73.5
  35. 35 gpt-5-2025-08-07-high-reasoning-high-reasoning 73.1
  36. 36 Claude Opus 4.5 73.1
  37. 37 GPT-5.4 Mini 73.0
  38. 38 GPT-5 72.8
  39. 39 Inkling 72.5
  40. 40 Gemini 3.5 Flash 71.8
  41. 41 GPT-5.2 71.6
  42. 42 Claude Sonnet 4.5 71.4
  43. 43 horizon-alpha 71.2
  44. 44 GLM-5 70.9
  45. 45 MiMo-V2-Pro 70.9
  46. 46 Muse Spark 1.1 70.6
  47. 47 Claude Opus 4.1 70.5
  48. 48 Kimi K2 Thinking 69.0
  49. 49 Gemini 3 Pro 68.8
  50. 50 Grok 4.20 68.5
  51. 51 Qwen3.5 397B-A17B 68.3
  52. 52 Gemini 3.1 Pro 68.2
  53. 53 Kimi K2 Jul 2025 68.1
  54. 54 GPT-5 Chat 66.8
  55. 55 DeepSeek-V3.2 66.8
  56. 56 Qwen3-Max 66.2
  57. 57 GLM-4.7 66.0
  58. 58 DeepSeek-V4-Flash 66.0
  59. 59 MiniMax-M2.5 65.3
  60. 60 Claude Haiku 4.5 65.0
  61. 61 Kimi K2 Sep 2025 63.8
  62. 62 hunter-alpha 62.7
  63. 63 o3 62.7
  64. 64 GPT-5.3 Instant 62.6
  65. 65 DeepSeek-V3.1 62.3
  66. 66 DeepSeek V4 Flash 0731 61.2
  67. 67 GPT-4o Mar 2025 60.2
  68. 68 DeepSeek-V3.2-Exp 60.2
  69. 69 DeepSeek-V3 Mar 2025 60.1
  70. 70 DeepSeek-R1 May 2025 59.5
  71. 71 tngtech/DeepSeek-R1T-Chimera 59.4
  72. 72 Qwen3.5-27B 59.0
  73. 73 Gemini 2.5 Pro Jun 2025 58.8
  74. 74 Claude 3.5 Sonnet 58.4
  75. 75 Claude Sonnet 4 58.4
  76. 76 MiniMax-M2.1 57.8
  77. 77 GPT-5.4 Nano 57.8
  78. 78 Claude 3.7 Sonnet 57.7
  79. 79 Qwen3-235B-A22B-Instruct Jul 2025 57.4
  80. 80 GLM-4.6 57.3
  81. 81 Gemma 4 31B IT 56.5
  82. 82 inclusionai/ling-1t 56.2
  83. 83 GPT-5 mini 56.2
  84. 84 Gemini 3 Flash 55.7
  85. 85 GLM-4.5 55.5
  86. 86 Qwen3.5-122B-A10B 54.3
  87. 87 Qwen 3.8 27B 53.3
  88. 88 MiMo-V2-Flash 53.1
  89. 89 AlexWortega/Qwen3.8-27B-antislop 52.4
  90. 90 Hermes 4 405B 51.2
  91. 91 healer-alpha 51.0
  92. 92 Gemma 4 26B A4B 50.7
  93. 93 Gemini 3.1 Flash-Lite 49.0
  94. 94 DeepSeek-V3.2-Speciale 48.8
  95. 95 Gemma 4 12B 48.7
  96. 96 Gemini 2.5 Flash Sep 2025 48.6
  97. 97 Mistral Large 3 48.3
  98. 98 glm-4.7-flash 47.8
  99. 99 bytedance-seed-1.6 47.3
  100. 100 openrouter/sherlock-dash-alpha 47.2
  101. 101 Olmo 3.1 32B Think 47.0
  102. 102 openrouter/bert-nebulon-alpha 46.4
  103. 103 MiniMax-M2 46.4
  104. 104 Mistral Medium 3.1 46.2
  105. 105 Grok 4 45.7
  106. 106 sam-paech/gemma-3-27b-it-antislop 45.2
  107. 107 Qwen3.5-35B-A3B 44.5
  108. 108 Mistral Small 4 43.2
  109. 109 sonoma-sky-alpha 42.9
  110. 110 Gemma 3 27B 41.9
  111. 111 Gemma 3 12B 41.8
  112. 112 Mistral Small 3.2 41.6
  113. 113 Reka Flash 3 41.2
  114. 114 Qwen3-32B 40.5
  115. 115 Grok 4.1 Fast 40.1
  116. 116 ftpo-exp501 39.7
  117. 117 Nova 2 Lite 38.8
  118. 118 QwQ-32B 36.9
  119. 119 Qwen3-14B 35.9
  120. 120 gpt-oss-120b 35.8
  121. 121 Qwen3-30B-A3B 35.6
  122. 122 LongCat-Flash-Chat 35.1
  123. 123 Trinity Large Preview 35.0
  124. 124 minimax-m2-her 34.8
  125. 125 Gemma 3 4B 34.4
  126. 126 nvidia/nemotron-3-nano-30b-a3b:free 32.9
  127. 127 mistral-small-creative 32.6
  128. 128 GPT-4.1 nano 32.4
  129. 129 Llama 4 Maverick 31.9
  130. 130 Ministral 3 14B 31.9
  131. 131 Nanbeige/Nanbeige4-3B-Thinking-2511 31.5
  132. 132 Qwen3-Next-80B-A3B-Instruct 31.5
  133. 133 Llama 4 Scout 31.1
  134. 134 openrouter/aurora-alpha 29.4
  135. 135 Mistral NeMo 28.7
  136. 136 GPT-5 nano 28.0
  137. 137 Solar Pro 3 27.0
  138. 138 Phi-4-multimodal 22.8
  139. 139 gpt-oss-20b 21.2
  140. 140 Qwen3-4B 17.0
#EntryScoreSettingDateSource
1 Claude Opus 5 86.3 Jul 2026 Official leaderboard
2 Claude Fable 5.1 85.3 Sep 2026 Official leaderboard
3 Claude Opus 5.5 84.6 Sep 2026 Official leaderboard
4 Claude Fable 5 83.0 Jun 2026 Official leaderboard
5 Grok 4.7 83.0 Sep 2026 Official leaderboard
6 Muse Spark 1.3 82.8 Sep 2026 Official leaderboard
7 GPT-6 Astra 82.8 Sep 2026 Official leaderboard
8 GPT-6 Sol 82.8 Sep 2026 Official leaderboard
9 aion-3.5 82.1 Official leaderboard
10 Claude Opus 4.7 81.8 Apr 2026 Official leaderboard
11 GLM-5.3 81.8 Aug 2026 Official leaderboard
12 GPT-5.6 Sol 81.7 Jul 2026 Official leaderboard
13 Muse Spark 1.2 81.5 Aug 2026 Official leaderboard
14 Gemini 4 Argon 80.9 Sep 2026 Official leaderboard
15 Claude Opus 4.8 80.8 May 2026 Official leaderboard
16 ox-alpha 79.9 Official leaderboard
17 Claude Sonnet 4.6 79.9 Feb 2026 Official leaderboard
18 Kimi K3.1 Pro 79.9 Oct 2026 Official leaderboard
19 Kimi K3 79.6 Jul 2026 Official leaderboard
20 Kimi K2.6 78.5 Apr 2026 Official leaderboard
21 GPT-5.4 78.3 Mar 2026 Official leaderboard
22 Claude Sonnet 5 78.3 Jun 2026 Official leaderboard
23 GPT-5.5 78.2 Apr 2026 Official leaderboard
24 GPT-5.6 Terra 78.0 Jul 2026 Official leaderboard
25 GLM-5.2 77.9 Jun 2026 Official leaderboard
26 Claude Opus 4.6 77.7 Feb 2026 Official leaderboard
27 Gemini 3.8 Flash 76.8 Sep 2026 Official leaderboard
28 DeepSeek-V4-Pro 75.6 Apr 2026 Official leaderboard
29 GPT-5.6 Luna 75.5 Jul 2026 Official leaderboard
30 Kimi K2.5 74.9 Jan 2026 Official leaderboard
31 Altworld/Hemmingway-1 74.2 Official leaderboard
32 DeepSeek V4.1 Flash 74.0 Sep 2026 Official leaderboard
33 MiMo-V2.5-Pro 73.7 Apr 2026 Official leaderboard
34 GLM-5.1 73.5 Apr 2026 Official leaderboard
35 gpt-5-2025-08-07-high-reasoning-high-reasoning 73.1 Official leaderboard
36 Claude Opus 4.5 73.1 Nov 2025 Official leaderboard
37 GPT-5.4 Mini 73.0 Mar 2026 Official leaderboard
38 GPT-5 72.8 Aug 2025 Official leaderboard
39 Inkling 72.5 Jul 2026 Official leaderboard
40 Gemini 3.5 Flash 71.8 May 2026 Official leaderboard
41 GPT-5.2 71.6 Dec 2025 Official leaderboard
42 Claude Sonnet 4.5 71.4 Sep 2025 Official leaderboard
43 horizon-alpha 71.2 Official leaderboard
44 GLM-5 70.9 Feb 2026 Official leaderboard
45 MiMo-V2-Pro 70.9 Mar 2026 Official leaderboard
46 Muse Spark 1.1 70.6 Jul 2026 Official leaderboard
47 Claude Opus 4.1 70.5 Aug 2025 Official leaderboard
48 Kimi K2 Thinking 69.0 Nov 2025 Official leaderboard
49 Gemini 3 Pro 68.8 Nov 2025 Official leaderboard
50 Grok 4.20 68.5 Feb 2026 Official leaderboard
51 Qwen3.5 397B-A17B 68.3 Feb 2026 Official leaderboard
52 Gemini 3.1 Pro 68.2 Feb 2026 Official leaderboard
53 Kimi K2 Jul 2025 68.1 Jul 2025 Official leaderboard
54 GPT-5 Chat 66.8 Aug 2025 Official leaderboard
55 DeepSeek-V3.2 66.8 Dec 2025 Official leaderboard
56 Qwen3-Max 66.2 Sep 2025 Official leaderboard
57 GLM-4.7 66.0 Dec 2025 Official leaderboard
58 DeepSeek-V4-Flash 66.0 Apr 2026 Official leaderboard
59 MiniMax-M2.5 65.3 Feb 2026 Official leaderboard
60 Claude Haiku 4.5 65.0 Oct 2025 Official leaderboard
61 Kimi K2 Sep 2025 63.8 Sep 2025 Official leaderboard
62 hunter-alpha 62.7 Official leaderboard
63 o3 62.7 Apr 2025 Official leaderboard
64 GPT-5.3 Instant 62.6 Mar 2026 Official leaderboard
65 DeepSeek-V3.1 62.3 Aug 2025 Official leaderboard
66 DeepSeek V4 Flash 0731 61.2 Jul 2026 Official leaderboard
67 GPT-4o Mar 2025 60.2 Mar 2025 Official leaderboard
68 DeepSeek-V3.2-Exp 60.2 Sep 2025 Official leaderboard
69 DeepSeek-V3 Mar 2025 60.1 Mar 2025 Official leaderboard
70 DeepSeek-R1 May 2025 59.5 May 2025 Official leaderboard
71 tngtech/DeepSeek-R1T-Chimera 59.4 Official leaderboard
72 Qwen3.5-27B 59.0 Feb 2026 Official leaderboard
73 Gemini 2.5 Pro Jun 2025 58.8 Jun 2025 Official leaderboard
74 Claude 3.5 Sonnet 58.4 Jun 2024 Official leaderboard
75 Claude Sonnet 4 58.4 May 2025 Official leaderboard
76 MiniMax-M2.1 57.8 Dec 2025 Official leaderboard
77 GPT-5.4 Nano 57.8 Mar 2026 Official leaderboard
78 Claude 3.7 Sonnet 57.7 Feb 2025 Official leaderboard
79 Qwen3-235B-A22B-Instruct Jul 2025 57.4 Jul 2025 Official leaderboard
80 GLM-4.6 57.3 Sep 2025 Official leaderboard
81 Gemma 4 31B IT 56.5 Apr 2026 Official leaderboard
82 inclusionai/ling-1t 56.2 Official leaderboard
83 GPT-5 mini 56.2 Aug 2025 Official leaderboard
84 Gemini 3 Flash 55.7 Dec 2025 Official leaderboard
85 GLM-4.5 55.5 Aug 2025 Official leaderboard
86 Qwen3.5-122B-A10B 54.3 Feb 2026 Official leaderboard
87 Qwen 3.8 27B 53.3 Aug 2026 Official leaderboard
88 MiMo-V2-Flash 53.1 Dec 2025 Official leaderboard
89 AlexWortega/Qwen3.8-27B-antislop 52.4 Official leaderboard
90 Hermes 4 405B 51.2 Aug 2025 Official leaderboard
91 healer-alpha 51.0 Official leaderboard
92 Gemma 4 26B A4B 50.7 Apr 2026 Official leaderboard
93 Gemini 3.1 Flash-Lite 49.0 Mar 2026 Official leaderboard
94 DeepSeek-V3.2-Speciale 48.8 Dec 2025 Official leaderboard
95 Gemma 4 12B 48.7 Jun 2026 Official leaderboard
96 Gemini 2.5 Flash Sep 2025 48.6 Sep 2025 Official leaderboard
97 Mistral Large 3 48.3 Dec 2025 Official leaderboard
98 glm-4.7-flash 47.8 Jan 2026 Official leaderboard
99 bytedance-seed-1.6 47.3 Official leaderboard
100 openrouter/sherlock-dash-alpha 47.2 Official leaderboard
101 Olmo 3.1 32B Think 47.0 Dec 2025 Official leaderboard
102 openrouter/bert-nebulon-alpha 46.4 Official leaderboard
103 MiniMax-M2 46.4 Oct 2025 Official leaderboard
104 Mistral Medium 3.1 46.2 Aug 2025 Official leaderboard
105 Grok 4 45.7 Jul 2025 Official leaderboard
106 sam-paech/gemma-3-27b-it-antislop 45.2 Official leaderboard
107 Qwen3.5-35B-A3B 44.5 Feb 2026 Official leaderboard
108 Mistral Small 4 43.2 Mar 2026 Official leaderboard
109 sonoma-sky-alpha 42.9 Official leaderboard
110 Gemma 3 27B 41.9 Mar 2025 Official leaderboard
111 Gemma 3 12B 41.8 Mar 2025 Official leaderboard
112 Mistral Small 3.2 41.6 Jun 2025 Official leaderboard
113 Reka Flash 3 41.2 Mar 2025 Official leaderboard
114 Qwen3-32B 40.5 Apr 2025 Official leaderboard
115 Grok 4.1 Fast 40.1 Nov 2025 Official leaderboard
116 ftpo-exp501 39.7 Official leaderboard
117 Nova 2 Lite 38.8 Dec 2025 Official leaderboard
118 QwQ-32B 36.9 Mar 2025 Official leaderboard
119 Qwen3-14B 35.9 Apr 2025 Official leaderboard
120 gpt-oss-120b 35.8 Aug 2025 Official leaderboard
121 Qwen3-30B-A3B 35.6 Apr 2025 Official leaderboard
122 LongCat-Flash-Chat 35.1 Aug 2025 Official leaderboard
123 Trinity Large Preview 35.0 Jan 2026 Official leaderboard
124 minimax-m2-her 34.8 Official leaderboard
125 Gemma 3 4B 34.4 Mar 2025 Official leaderboard
126 nvidia/nemotron-3-nano-30b-a3b:free 32.9 Official leaderboard
127 mistral-small-creative 32.6 Official leaderboard
128 GPT-4.1 nano 32.4 Apr 2025 Official leaderboard
129 Llama 4 Maverick 31.9 Apr 2025 Official leaderboard
130 Ministral 3 14B 31.9 Dec 2025 Official leaderboard
131 Nanbeige/Nanbeige4-3B-Thinking-2511 31.5 Official leaderboard
132 Qwen3-Next-80B-A3B-Instruct 31.5 Sep 2025 Official leaderboard
133 Llama 4 Scout 31.1 Apr 2025 Official leaderboard
134 openrouter/aurora-alpha 29.4 Official leaderboard
135 Mistral NeMo 28.7 Jul 2024 Official leaderboard
136 GPT-5 nano 28.0 Aug 2025 Official leaderboard
137 Solar Pro 3 27.0 Jan 2026 Official leaderboard
138 Phi-4-multimodal 22.8 Feb 2025 Official leaderboard
139 gpt-oss-20b 21.2 Aug 2025 Official leaderboard
140 Qwen3-4B 17.0 Apr 2025 Official leaderboard

11 models have held number one since its release in May 2025

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

OpenAI Moonshot AI Anthropic
Own leaderboard 11 leaders
Claude Opus 5 86.3

Scores over time

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

On the benchmark’s own leaderboard
0.00 20.0 40.0 60.0 80.0 100.0 2024 2025 2026 2027 0.00 20.0 40.0 60.0 80.0 100.0 2024 2025 2026 2027

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Writing and taste

All 24 in Writing and taste
  • Design Bench (Moonshot) In use Models make graphic designs from direct prompts or few-shot reference images. An automated grader rates the results as Elo scores. Kimi K3.1 Pro 957
  • WritingBench Fading Models complete 1,000 real-world writing requests across 6 domains and 100 subdomains. A judge model scores each reply on five query-specific criteria. GPT-5 83.9
  • Creative Writing (MiniMax in-house) Only in system cards Models write creative pieces for real Hailuo AI user requests. Automatic assessments score the writing. MiniMax-Text-01 81.3%
  • DeepSeek Chinese functional writing evaluation Only in system cards Models answer 3,170 everyday Chinese writing prompts such as reports, emails and speeches. Pairwise judging against a baseline model gives a win rate. DeepSeek-V4-Pro 62.6%
  • Slide generation human evaluation (Z.ai) Only in system cards People compare HTML slide decks from two models on content, layout and visual quality. Metric is overall win rate. GLM-5 67.5%
  • Task-dependent output homogenization evaluation Only in system cards Models answer prompts several times. A judge scores functional diversity where it is wanted and consistency where one answer is right. Muse Spark 0.21