Benchmark Almanac

StudyArena Text

LongCat-2.0 1,375 leads since July 2026
Not in model cards

Models produce text responses to user prompts. Anonymous human comparisons determine the published rating in this separate vote pool.

Only entries with observed ballots count as results. Topic filters do not form separate benchmark records.

Made by
StudyArena
Kind
Overall chat quality
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Strongest model tested
GPT-6 Astra #3 on the capability index
Last model added
October 2026
Scores refreshed
November 12th, 2026
106
Entries on its own leaderboard, 85 models we track

LongCat-2.0 leads with 1,375

Best score per entry from the benchmark's own leaderboard, read on November 12th, 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.

106 entries. Scroll sideways for the rest. Bars start at 1,000.

  1. 1,375 LongCat-2.0
  2. 1,347 MiniMax-M3
  3. 1,347 Muse Spark 1.3
  4. 1,339 Gemini 3.7 Flash
  5. 1,312 GLM-5.2
  6. 1,307 Qwen3.7-Plus
  7. 1,286 GPT-5.6 Luna
  8. 1,283 Qwen3.8-2.4T-A95B
  9. 1,282 Muse Spark 1.1
  10. 1,281 Nemotron 3 Nano Omni 30B-A3B
  11. 1,280 Claude Fable 5
  12. 1,276 GPT-5.6 Terra
  13. 1,274 Muse Wonder
  14. 1,270 MiMo-V2.5-Pro
  15. 1,270 Ling-3.0-flash
  16. 1,268 Claude Opus 5.5
  17. 1,265 Gemini 3.8 Flash
  18. 1,263 GPT-5.6 Sol
  19. 1,263 Claude Fable 5.1
  20. 1,259 Muse Spark 1.2
  21. 1,257 GLM-5.3
  22. 1,255 Gemini 3.6 Flash
  23. 1,250 Grok 4.20
  24. 1,250 Qwen3.7 Flash
  25. 1,249 Solar Pro 4
  26. 1,249 GPT-6 Astra
  27. 1,247 Gemini 4 Argon
  28. 1,244 Kimi K3
  29. 1,243 Aion 3.0
  30. 1,243 Seed 2.1 Turbo
  31. 1,243 DeepSeek V4 Pro 0813
  32. 1,243 GPT-6 Luna
  33. 1,242 DeepSeek-V4-Flash
  34. 1,242 Claude Opus 5
  35. 1,241 Claude Sonnet 5
  36. 1,240 Grok 4.5
  37. 1,238 Tencent Hy3
  38. 1,238 Mercury 2.5
  39. 1,236 Mistral Small 3
  40. 1,236 Laguna XS 2.1
  41. 1,236 Gemini 3.5 Flash-Lite
  42. 1,236 GPT-6.1 Sol
  43. 1,235 Gemini 3.1 Flash-Lite
  44. 1,234 Gemini 3.1 Pro
  45. 1,234 DeepSeek V4.1 Flash
  46. 1,231 Sonar Reasoning Pro
  47. 1,229 Sonar Pro
  48. 1,229 Nemotron 3 Ultra
  49. 1,229 Tencent Hy4 preview
  50. 1,228 sonar
  51. 1,227 Inkling-Small
  52. 1,227 GLM-5.3-Flash
  53. 1,225 Kimi K2.6
  54. 1,223 GPT-6 Sol
  55. 1,222 Mercury 2
  56. 1,222 Grok 4.7
  57. 1,221 GLM-4.7
  58. 1,220 inclusionai/ling-3.0-flash:free
  59. 1,219 Claude Haiku 4.5
  60. 1,219 MiMo-V2.5
  61. 1,219 Qwen 3.8 27B
  62. 1,217 Solar Pro 3
  63. 1,214 Mistral Medium 3.1
  64. 1,212 Seed 2.0 Code
  65. 1,212 Qwen3.8 Max (0902)
  66. 1,211 MiMo-V2.6 Pro
  67. 1,210 Grok 4.6
  68. 1,205 Aion 2.0
  69. 1,204 Aion 3.0 Mini
  70. 1,200 Union Alpha
  71. 1,199 MiMo-V2.6 Flash
  72. 1,197 Qwen3.8 Flash
  73. 1,196 gpt-oss-120b
  74. 1,193 Laguna S 2.1
  75. 1,192 Sabiá 4
  76. 1,191 Nova 2 Lite
  77. 1,190 Inkling
  78. 1,189 Muse Glimmer 30B
  79. 1,189 MiniMax-M2.5
  80. 1,188 Gemma 4 31B IT
  81. 1,188 Mistral Medium 3.5
  82. 1,187 Grok 4.3
  83. 1,186 Seed2.0 Mini
  84. 1,184 Llama 4 Maverick
  85. 1,183 Parallel Speed
  86. 1,183 gpt-oss-20b
  87. 1,180 Nemotron 3.5 Lightning 30B-A3B
  88. 1,176 Parallel Base
  89. 1,174 glm-4.7-flash
  90. 1,173 Granite 4.2 8B
  91. 1,170 Claude Sonnet 5.5
  92. 1,167 Nemotron 3 Super
  93. 1,166 Nemotron 3 Nano 30B
  94. 1,166 Parallel Lite
  95. 1,164 Sabiazinho 4
  96. 1,159 Llama 4 Scout
  97. 1,157 Granite 4.1 8B
  98. 1,156 Nova Lite
  99. 1,151 Aion-RP 1.0 (8B)
  100. 1,148 granite-4.0-micro
  101. 1,131 Nemotron Nano 9B
  102. 1,131 Step 5 Preview
  103. 1,124 Step 3.5 Flash
  104. 1,089 Nemotron Nano 12B VL
  105. 1,060 Sonar Deep Research
  106. 1,026 deepseek-chat

Bars start at 1,000.

  1. 1 LongCat-2.0 high reasoning 1,375
  2. 2 MiniMax-M3 medium reasoning 1,347
  3. 3 Muse Spark 1.3 xhigh reasoning 1,347
  4. 4 Gemini 3.7 Flash minimal reasoning 1,339
  5. 5 GLM-5.2 medium reasoning 1,312
  6. 6 Qwen3.7-Plus 1,307
  7. 7 GPT-5.6 Luna high reasoning 1,286
  8. 8 Qwen3.8-2.4T-A95B high reasoning 1,283
  9. 9 Muse Spark 1.1 high reasoning 1,282
  10. 10 Nemotron 3 Nano Omni 30B-A3B 1,281
  11. 11 Claude Fable 5 high reasoning 1,280
  12. 12 GPT-5.6 Terra low reasoning 1,276
  13. 13 Muse Wonder xhigh reasoning 1,274
  14. 14 MiMo-V2.5-Pro low reasoning 1,270
  15. 15 Ling-3.0-flash low reasoning 1,270
  16. 16 Claude Opus 5.5 medium reasoning 1,268
  17. 17 Gemini 3.8 Flash medium reasoning 1,265
  18. 18 GPT-5.6 Sol medium reasoning 1,263
  19. 19 Claude Fable 5.1 medium reasoning 1,263
  20. 20 Muse Spark 1.2 high reasoning 1,259
  21. 21 GLM-5.3 high reasoning 1,257
  22. 22 Gemini 3.6 Flash minimal reasoning 1,255
  23. 23 Grok 4.20 1,250
  24. 24 Qwen3.7 Flash 1,250
  25. 25 Solar Pro 4 reasoning 1,249
  26. 26 GPT-6 Astra high reasoning 1,249
  27. 27 Gemini 4 Argon medium reasoning 1,247
  28. 28 Kimi K3 high reasoning 1,244
  29. 29 Aion 3.0 reasoning 1,243
  30. 30 Seed 2.1 Turbo 1,243
  31. 31 DeepSeek V4 Pro 0813 1,243
  32. 32 GPT-6 Luna xhigh reasoning 1,243
  33. 33 DeepSeek-V4-Flash low reasoning 1,242
  34. 34 Claude Opus 5 medium reasoning 1,242
  35. 35 Claude Sonnet 5 low reasoning 1,241
  36. 36 Grok 4.5 low reasoning 1,240
  37. 37 Tencent Hy3 medium reasoning 1,238
  38. 38 Mercury 2.5 1,238
  39. 39 Mistral Small 3 high reasoning 1,236
  40. 40 Laguna XS 2.1 reasoning 1,236
  41. 41 Gemini 3.5 Flash-Lite minimal reasoning 1,236
  42. 42 GPT-6.1 Sol high reasoning 1,236
  43. 43 Gemini 3.1 Flash-Lite low reasoning 1,235
  44. 44 Gemini 3.1 Pro low reasoning 1,234
  45. 45 DeepSeek V4.1 Flash high reasoning 1,234
  46. 46 Sonar Reasoning Pro 1,231
  47. 47 Sonar Pro 1,229
  48. 48 Nemotron 3 Ultra 1,229
  49. 49 Tencent Hy4 preview 1,229
  50. 50 sonar 1,228
  51. 51 Inkling-Small high reasoning 1,227
  52. 52 GLM-5.3-Flash high reasoning 1,227
  53. 53 Kimi K2.6 1,225
  54. 54 GPT-6 Sol medium reasoning 1,223
  55. 55 Mercury 2 medium reasoning 1,222
  56. 56 Grok 4.7 medium reasoning 1,222
  57. 57 GLM-4.7 reasoning 1,221
  58. 58 inclusionai/ling-3.0-flash:free 1,220
  59. 59 Claude Haiku 4.5 1,219
  60. 60 MiMo-V2.5 1,219
  61. 61 Qwen 3.8 27B medium reasoning 1,219
  62. 62 Solar Pro 3 high reasoning 1,217
  63. 63 Mistral Medium 3.1 1,214
  64. 64 Seed 2.0 Code high reasoning 1,212
  65. 65 Qwen3.8 Max (0902) minimal reasoning 1,212
  66. 66 MiMo-V2.6 Pro 1,211
  67. 67 Grok 4.6 low reasoning 1,210
  68. 68 Aion 2.0 reasoning 1,205
  69. 69 Aion 3.0 Mini reasoning 1,204
  70. 70 Union Alpha 1,200
  71. 71 MiMo-V2.6 Flash 1,199
  72. 72 Qwen3.8 Flash 1,197
  73. 73 gpt-oss-120b medium reasoning 1,196
  74. 74 Laguna S 2.1 1,193
  75. 75 Sabiá 4 1,192
  76. 76 Nova 2 Lite reasoning 1,191
  77. 77 Inkling high reasoning 1,190
  78. 78 Muse Glimmer 30B medium reasoning 1,189
  79. 79 MiniMax-M2.5 reasoning 1,189
  80. 80 Gemma 4 31B IT 1,188
  81. 81 Mistral Medium 3.5 1,188
  82. 82 Grok 4.3 low reasoning 1,187
  83. 83 Seed2.0 Mini medium reasoning 1,186
  84. 84 Llama 4 Maverick 1,184
  85. 85 Parallel Speed 1,183
  86. 86 gpt-oss-20b high reasoning 1,183
  87. 87 Nemotron 3.5 Lightning 30B-A3B 1,180
  88. 88 Parallel Base 1,176
  89. 89 glm-4.7-flash reasoning 1,174
  90. 90 Granite 4.2 8B low reasoning 1,173
  91. 91 Claude Sonnet 5.5 high reasoning 1,170
  92. 92 Nemotron 3 Super 1,167
  93. 93 Nemotron 3 Nano 30B reasoning 1,166
  94. 94 Parallel Lite 1,166
  95. 95 Sabiazinho 4 1,164
  96. 96 Llama 4 Scout 1,159
  97. 97 Granite 4.1 8B 1,157
  98. 98 Nova Lite 1,156
  99. 99 Aion-RP 1.0 (8B) 1,151
  100. 100 granite-4.0-micro 1,148
  101. 101 Nemotron Nano 9B reasoning 1,131
  102. 102 Step 5 Preview high reasoning 1,131
  103. 103 Step 3.5 Flash 1,124
  104. 104 Nemotron Nano 12B VL reasoning 1,089
  105. 105 Sonar Deep Research medium reasoning 1,060
  106. 106 deepseek-chat 1,026

Source: Official StudyArena Text leaderboard

#EntryScoreSettingDateSource
1 LongCat-2.0 1,375 high reasoning Jul 2026 Official leaderboard
2 MiniMax-M3 1,347 medium reasoning Jun 2026 Official leaderboard
3 Muse Spark 1.3 1,347 xhigh reasoning Sep 2026 Official leaderboard
4 Gemini 3.7 Flash 1,339 minimal reasoning Aug 2026 Official leaderboard
5 GLM-5.2 1,312 medium reasoning Jun 2026 Official leaderboard
6 Qwen3.7-Plus 1,307 Jun 2026 Official leaderboard
7 GPT-5.6 Luna 1,286 high reasoning Jul 2026 Official leaderboard
8 Qwen3.8-2.4T-A95B 1,283 high reasoning Aug 2026 Official leaderboard
9 Muse Spark 1.1 1,282 high reasoning Jul 2026 Official leaderboard
10 Nemotron 3 Nano Omni 30B-A3B 1,281 Apr 2026 Official leaderboard
11 Claude Fable 5 1,280 high reasoning Jun 2026 Official leaderboard
12 GPT-5.6 Terra 1,276 low reasoning Jul 2026 Official leaderboard
13 Muse Wonder 1,274 xhigh reasoning Oct 2026 Official leaderboard
14 MiMo-V2.5-Pro 1,270 low reasoning Apr 2026 Official leaderboard
15 Ling-3.0-flash 1,270 low reasoning Aug 2026 Official leaderboard
16 Claude Opus 5.5 1,268 medium reasoning Sep 2026 Official leaderboard
17 Gemini 3.8 Flash 1,265 medium reasoning Sep 2026 Official leaderboard
18 GPT-5.6 Sol 1,263 medium reasoning Jul 2026 Official leaderboard
19 Claude Fable 5.1 1,263 medium reasoning Sep 2026 Official leaderboard
20 Muse Spark 1.2 1,259 high reasoning Aug 2026 Official leaderboard
21 GLM-5.3 1,257 high reasoning Aug 2026 Official leaderboard
22 Gemini 3.6 Flash 1,255 minimal reasoning Jul 2026 Official leaderboard
23 Grok 4.20 1,250 Feb 2026 Official leaderboard
24 Qwen3.7 Flash 1,250 Jul 2026 Official leaderboard
25 Solar Pro 4 1,249 reasoning Aug 2026 Official leaderboard
26 GPT-6 Astra 1,249 high reasoning Sep 2026 Official leaderboard
27 Gemini 4 Argon 1,247 medium reasoning Sep 2026 Official leaderboard
28 Kimi K3 1,244 high reasoning Jul 2026 Official leaderboard
29 Aion 3.0 1,243 reasoning Official leaderboard
30 Seed 2.1 Turbo 1,243 Official leaderboard
31 DeepSeek V4 Pro 0813 1,243 Aug 2026 Official leaderboard
32 GPT-6 Luna 1,243 xhigh reasoning Sep 2026 Official leaderboard
33 DeepSeek-V4-Flash 1,242 low reasoning Apr 2026 Official leaderboard
34 Claude Opus 5 1,242 medium reasoning Jul 2026 Official leaderboard
35 Claude Sonnet 5 1,241 low reasoning Jun 2026 Official leaderboard
36 Grok 4.5 1,240 low reasoning Jul 2026 Official leaderboard
37 Tencent Hy3 1,238 medium reasoning Jul 2026 Official leaderboard
38 Mercury 2.5 1,238 Sep 2026 Official leaderboard
39 Mistral Small 3 1,236 high reasoning Jan 2025 Official leaderboard
40 Laguna XS 2.1 1,236 reasoning Jul 2026 Official leaderboard
41 Gemini 3.5 Flash-Lite 1,236 minimal reasoning Jul 2026 Official leaderboard
42 GPT-6.1 Sol 1,236 high reasoning Sep 2026 Official leaderboard
43 Gemini 3.1 Flash-Lite 1,235 low reasoning Mar 2026 Official leaderboard
44 Gemini 3.1 Pro 1,234 low reasoning Feb 2026 Official leaderboard
45 DeepSeek V4.1 Flash 1,234 high reasoning Sep 2026 Official leaderboard
46 Sonar Reasoning Pro 1,231 Official leaderboard
47 Sonar Pro 1,229 Jan 2025 Official leaderboard
48 Nemotron 3 Ultra 1,229 Jun 2026 Official leaderboard
49 Tencent Hy4 preview 1,229 Aug 2026 Official leaderboard
50 sonar 1,228 Jan 2025 Official leaderboard
51 Inkling-Small 1,227 high reasoning Jul 2026 Official leaderboard
52 GLM-5.3-Flash 1,227 high reasoning Aug 2026 Official leaderboard
53 Kimi K2.6 1,225 Apr 2026 Official leaderboard
54 GPT-6 Sol 1,223 medium reasoning Sep 2026 Official leaderboard
55 Mercury 2 1,222 medium reasoning Feb 2026 Official leaderboard
56 Grok 4.7 1,222 medium reasoning Sep 2026 Official leaderboard
57 GLM-4.7 1,221 reasoning Dec 2025 Official leaderboard
58 inclusionai/ling-3.0-flash:free 1,220 Official leaderboard
59 Claude Haiku 4.5 1,219 Oct 2025 Official leaderboard
60 MiMo-V2.5 1,219 Apr 2026 Official leaderboard
61 Qwen 3.8 27B 1,219 medium reasoning Aug 2026 Official leaderboard
62 Solar Pro 3 1,217 high reasoning Jan 2026 Official leaderboard
63 Mistral Medium 3.1 1,214 Aug 2025 Official leaderboard
64 Seed 2.0 Code 1,212 high reasoning Official leaderboard
65 Qwen3.8 Max (0902) 1,212 minimal reasoning Sep 2026 Official leaderboard
66 MiMo-V2.6 Pro 1,211 Sep 2026 Official leaderboard
67 Grok 4.6 1,210 low reasoning Aug 2026 Official leaderboard
68 Aion 2.0 1,205 reasoning Official leaderboard
69 Aion 3.0 Mini 1,204 reasoning Official leaderboard
70 Union Alpha 1,200 Official leaderboard
71 MiMo-V2.6 Flash 1,199 Sep 2026 Official leaderboard
72 Qwen3.8 Flash 1,197 Aug 2026 Official leaderboard
73 gpt-oss-120b 1,196 medium reasoning Aug 2025 Official leaderboard
74 Laguna S 2.1 1,193 Jul 2026 Official leaderboard
75 Sabiá 4 1,192 Official leaderboard
76 Nova 2 Lite 1,191 reasoning Dec 2025 Official leaderboard
77 Inkling 1,190 high reasoning Jul 2026 Official leaderboard
78 Muse Glimmer 30B 1,189 medium reasoning Official leaderboard
79 MiniMax-M2.5 1,189 reasoning Feb 2026 Official leaderboard
80 Gemma 4 31B IT 1,188 Apr 2026 Official leaderboard
81 Mistral Medium 3.5 1,188 Apr 2026 Official leaderboard
82 Grok 4.3 1,187 low reasoning Apr 2026 Official leaderboard
83 Seed2.0 Mini 1,186 medium reasoning Feb 2026 Official leaderboard
84 Llama 4 Maverick 1,184 Apr 2025 Official leaderboard
85 Parallel Speed 1,183 Official leaderboard
86 gpt-oss-20b 1,183 high reasoning Aug 2025 Official leaderboard
87 Nemotron 3.5 Lightning 30B-A3B 1,180 Aug 2026 Official leaderboard
88 Parallel Base 1,176 Official leaderboard
89 glm-4.7-flash 1,174 reasoning Jan 2026 Official leaderboard
90 Granite 4.2 8B 1,173 low reasoning Aug 2026 Official leaderboard
91 Claude Sonnet 5.5 1,170 high reasoning Sep 2026 Official leaderboard
92 Nemotron 3 Super 1,167 Mar 2026 Official leaderboard
93 Nemotron 3 Nano 30B 1,166 reasoning Official leaderboard
94 Parallel Lite 1,166 Official leaderboard
95 Sabiazinho 4 1,164 Official leaderboard
96 Llama 4 Scout 1,159 Apr 2025 Official leaderboard
97 Granite 4.1 8B 1,157 Apr 2026 Official leaderboard
98 Nova Lite 1,156 Official leaderboard
99 Aion-RP 1.0 (8B) 1,151 Official leaderboard
100 granite-4.0-micro 1,148 Oct 2025 Official leaderboard
101 Nemotron Nano 9B 1,131 reasoning Official leaderboard
102 Step 5 Preview 1,131 high reasoning Sep 2026 Official leaderboard
103 Step 3.5 Flash 1,124 Feb 2026 Official leaderboard
104 Nemotron Nano 12B VL 1,089 reasoning Official leaderboard
105 Sonar Deep Research 1,060 medium reasoning Official leaderboard
106 deepseek-chat 1,026 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
LongCat-2.0
reasoning low reasoni… medium reas… high reason…
1,375
MiniMax-M3
reasoning medium reas… high reason…
1,347
Muse Spark 1.3
high reason… xhigh reaso…
1,347
Gemini 3.7 Flash
minimal rea… low reasoni…
1,339
GLM-5.2
low reasoni… medium reas… high reason…
1,312
Qwen3.7-Plus
reasoning
1,307
GPT-5.6 Luna
low reasoni…
1,286
Qwen3.8-2.4T-A95B
minimal rea… high reason… xhigh reaso…
1,283
Muse Spark 1.1
minimal rea… medium reas… high reason… xhigh reaso…
1,282
Nemotron 3 Nano Omni 30B-A3B
reasoning
1,281
Claude Fable 5
low reasoni…
1,280
GPT-5.6 Terra
low reasoni…
1,276

Source: Official StudyArena Text leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
LongCat-2.0 high reasoning 1,375
MiniMax-M3 medium reasoning 1,347
Muse Spark 1.3 xhigh reasoning 1,347
Gemini 3.7 Flash minimal reasoning 1,339
GLM-5.2 medium reasoning 1,312
Qwen3.7-Plus 1,307
GPT-5.6 Luna high reasoning 1,286
Qwen3.8-2.4T-A95B high reasoning 1,283
Muse Spark 1.1 high reasoning 1,282
Nemotron 3 Nano Omni 30B-A3B 1,281
Claude Fable 5 high reasoning 1,280
GPT-5.6 Terra low reasoning 1,276
Muse Wonder xhigh reasoning 1,274
GPT-5.6 Terra 1,273
GPT-5.6 Luna low reasoning 1,272
MiMo-V2.5-Pro low reasoning 1,270
Ling-3.0-flash low reasoning 1,270
Claude Opus 5.5 medium reasoning 1,268
Muse Wonder low reasoning 1,268
Claude Fable 5 low reasoning 1,266
Gemini 3.8 Flash medium reasoning 1,265
Muse Wonder medium reasoning 1,265
GPT-5.6 Sol medium reasoning 1,263
Claude Fable 5.1 medium reasoning 1,263
Claude Opus 5.5 high reasoning 1,261
Muse Spark 1.2 high reasoning 1,259
GPT-5.6 Sol 1,258
Ling-3.0-flash high reasoning 1,258
GLM-5.3 high reasoning 1,257
Gemini 3.8 Flash low reasoning 1,257
GPT-5.6 Sol low reasoning 1,256
Claude Fable 5.1 high reasoning 1,256
GPT-5.6 Sol high reasoning 1,255
Gemini 3.6 Flash minimal reasoning 1,255
GLM-5.2 1,254
Claude Fable 5.1 max reasoning 1,254
Grok 4.20 1,250
Qwen3.7 Flash 1,250
Qwen3.8-2.4T-A95B medium reasoning 1,250
GPT-5.6 Terra high reasoning 1,249
Muse Spark 1.2 minimal reasoning 1,249
Solar Pro 4 reasoning 1,249
GPT-6 Astra high reasoning 1,249
Muse Wonder minimal reasoning 1,249
Claude Fable 5 medium reasoning 1,247
Gemini 4 Argon medium reasoning 1,247
Gemini 3.6 Flash high reasoning 1,246
Muse Spark 1.2 xhigh reasoning 1,246
Qwen3.7-Plus reasoning 1,244
GPT-5.6 Terra medium reasoning 1,244
Kimi K3 high reasoning 1,244
Claude Fable 5.1 low reasoning 1,244
Claude Opus 5.5 low reasoning 1,244
Muse Spark 1.1 xhigh reasoning 1,243
DeepSeek V4 Pro 0813 1,243
GPT-6 Luna xhigh reasoning 1,243
DeepSeek-V4-Flash low reasoning 1,242
MiniMax-M3 low reasoning 1,242
Claude Opus 5 medium reasoning 1,242
Gemini 4 Argon high reasoning 1,242
Claude Sonnet 5 low reasoning 1,241
GPT-6 Astra low reasoning 1,241
GPT-6 Luna medium reasoning 1,241
Grok 4.5 low reasoning 1,240
Gemini 3.7 Flash high reasoning 1,240
MiniMax-M3 reasoning 1,239
GPT-5.6 Luna medium reasoning 1,239
Tencent Hy3 medium reasoning 1,238
Grok 4.5 medium reasoning 1,238
Gemini 3.6 Flash medium reasoning 1,238
Claude Opus 5 high reasoning 1,238
Ling-3.0-flash 1,238
DeepSeek V4 Pro 0813 low reasoning 1,238
Gemini 3.7 Flash medium reasoning 1,238
Mercury 2.5 1,238
GLM-5.2 xhigh reasoning 1,237
Gemini 4 Argon low reasoning 1,237
Mistral Small 3 high reasoning 1,236
Laguna XS 2.1 reasoning 1,236
Muse Spark 1.1 medium reasoning 1,236
Gemini 3.5 Flash-Lite minimal reasoning 1,236
Muse Spark 1.2 medium reasoning 1,236
DeepSeek V4 Pro 0813 medium reasoning 1,236
GPT-6.1 Sol high reasoning 1,236
Gemini 3.1 Flash-Lite low reasoning 1,235
GPT-5.6 Luna 1,235
Gemini 3.7 Flash low reasoning 1,235
Gemini 3.1 Pro low reasoning 1,234
Gemini 3.8 Flash high reasoning 1,234
DeepSeek V4.1 Flash high reasoning 1,234
Grok 4.5 high reasoning 1,233
Claude Opus 5 low reasoning 1,233
DeepSeek V4 Pro 0813 high reasoning 1,233
GLM-5.3 low reasoning 1,233
Kimi K3 low reasoning 1,232
DeepSeek-V4-Flash 1,231
Gemini 3.5 Flash-Lite medium reasoning 1,231
MiMo-V2.5-Pro reasoning 1,230
Kimi K3 1,230
Qwen3.8-2.4T-A95B low reasoning 1,230
Sonar Pro 1,229
Nemotron 3 Ultra 1,229
Tencent Hy4 preview 1,229
sonar 1,228
GPT-6 Astra medium reasoning 1,228
Claude Opus 5.5 xhigh reasoning 1,228
Inkling-Small high reasoning 1,227
Ling-3.0-flash medium reasoning 1,227
GLM-5.3-Flash high reasoning 1,227
GPT-6 Luna high reasoning 1,227
GLM-5.3-Flash low reasoning 1,226
Gemini 3.1 Pro medium reasoning 1,225
Kimi K2.6 1,225
Ling-3.0-flash reasoning 1,225
Nemotron 3 Ultra high reasoning 1,224
GLM-5.2 high reasoning 1,224
Muse Wonder high reasoning 1,224
GPT-6 Sol medium reasoning 1,223
Mercury 2 medium reasoning 1,222
Qwen3.8-2.4T-A95B minimal reasoning 1,222
Grok 4.7 medium reasoning 1,222
GLM-4.7 reasoning 1,221
MiniMax-M3 1,220
Gemini 3.6 Flash low reasoning 1,220
DeepSeek V4.1 Flash low reasoning 1,220
Claude Haiku 4.5 1,219
MiMo-V2.5 1,219
Qwen 3.8 27B medium reasoning 1,219
GPT-6 Luna low reasoning 1,219
Muse Wonder max reasoning 1,219
Gemini 3.1 Flash-Lite high reasoning 1,218
DeepSeek-V4-Flash medium reasoning 1,218
DeepSeek V4.1 Flash 1,218
Solar Pro 3 high reasoning 1,217
Mercury 2 high reasoning 1,217
GLM-4.7 1,216
Gemini 3.5 Flash-Lite high reasoning 1,216
Muse Spark 1.2 low reasoning 1,216
Grok 4.7 low reasoning 1,216
GPT-6 Sol low reasoning 1,215
Mistral Medium 3.1 1,214
Gemini 3.1 Flash-Lite minimal reasoning 1,213
DeepSeek-V4-Flash high reasoning 1,213
Gemini 3.1 Pro high reasoning 1,212
Qwen3.7 Flash reasoning 1,212
Qwen3.8 Max (0902) minimal reasoning 1,212
Kimi K2.6 reasoning 1,211
MiMo-V2.6 Pro 1,211
Grok 4.6 low reasoning 1,210
Kimi K3 max reasoning 1,209
LongCat-2.0 1,208
LongCat-2.0 reasoning 1,208
GLM-5.3 max reasoning 1,208
Claude Sonnet 5 medium reasoning 1,207
Nemotron 3 Ultra medium reasoning 1,204
GPT-6 Luna max reasoning 1,204
GPT-6 Sol max reasoning 1,203
Gemini 3.5 Flash-Lite low reasoning 1,201
GPT-6 Astra xhigh reasoning 1,201
Claude Sonnet 5 high reasoning 1,200
Solar Pro 4 1,200
Qwen 3.8 27B low reasoning 1,200
Gemini 3.1 Flash-Lite medium reasoning 1,199
MiMo-V2.6 Flash 1,199
Claude Fable 5.1 xhigh reasoning 1,198
Mercury 2.5 medium reasoning 1,198
Qwen3.8 Flash 1,197
Qwen3.8 Max (0902) low reasoning 1,197
gpt-oss-120b medium reasoning 1,196
Qwen3.8-2.4T-A95B xhigh reasoning 1,196
GPT-6 Sol high reasoning 1,196
GLM-5.3-Flash max reasoning 1,194
Laguna S 2.1 1,193
Tencent Hy4 preview low reasoning 1,193
Nova 2 Lite reasoning 1,191
Nova 2 Lite 1,191
GLM-5.2 low reasoning 1,190
Inkling high reasoning 1,190
MiniMax-M2.5 reasoning 1,189
Qwen 3.8 27B 1,189
DeepSeek V4.1 Flash max reasoning 1,189
Gemma 4 31B IT 1,188
Mistral Medium 3.5 1,188
Mistral Medium 3.5 high reasoning 1,187
Grok 4.3 low reasoning 1,187
Grok 4.3 high reasoning 1,187
Inkling max reasoning 1,187
MiMo-V2.6 Flash reasoning 1,187
Seed2.0 Mini medium reasoning 1,186
MiMo-V2.5-Pro 1,186
GPT-6 Astra max reasoning 1,186
Gemma 4 31B IT reasoning 1,185
Llama 4 Maverick 1,184
Tencent Hy3 low reasoning 1,184
gpt-oss-20b high reasoning 1,183
Inkling medium reasoning 1,183
Grok 4.6 high reasoning 1,181
Nemotron 3.5 Lightning 30B-A3B 1,180
Solar Pro 3 1,179
Seed2.0 Mini 1,179
Seed2.0 Mini low reasoning 1,179
Laguna S 2.1 reasoning 1,179
Seed2.0 Mini minimal reasoning 1,177
Inkling low reasoning 1,175
glm-4.7-flash reasoning 1,174
Qwen 3.8 27B xhigh reasoning 1,174
Qwen3.8 Max (0902) medium reasoning 1,174
Seed2.0 Mini high reasoning 1,173
Granite 4.2 8B low reasoning 1,173
GPT-6 Sol xhigh reasoning 1,173
gpt-oss-20b low reasoning 1,172
Solar Pro 3 reasoning 1,172
Inkling-Small low reasoning 1,172
Inkling-Small max reasoning 1,172
Mercury 2.5 low reasoning 1,172
gpt-oss-120b low reasoning 1,171
MiMo-V2.5-Pro medium reasoning 1,170
Claude Sonnet 5.5 high reasoning 1,170
Nemotron 3.5 Lightning 30B-A3B reasoning 1,169
Nemotron 3 Super 1,167
Tencent Hy4 preview high reasoning 1,167
glm-4.7-flash 1,166
LongCat-2.0 low reasoning 1,166
Grok 4.3 medium reasoning 1,164
Grok 4.7 high reasoning 1,163
Granite 4.2 8B 1,162
MiMo-V2.5 reasoning 1,160
Inkling-Small medium reasoning 1,160
Llama 4 Scout 1,159
Laguna XS 2.1 1,159
Tencent Hy3 high reasoning 1,158
Grok 4.6 medium reasoning 1,158
Granite 4.1 8B 1,157
Muse Spark 1.1 low reasoning 1,153
Mercury 2.5 high reasoning 1,152
Qwen3.8 Flash reasoning 1,151
Inkling-Small 1,150
Tencent Hy3 1,149
Inkling minimal reasoning 1,149
granite-4.0-micro 1,148
Grok 4.3 1,147
gpt-oss-20b medium reasoning 1,146
Mistral Medium 3.5 low reasoning 1,146
MiMo-V2.6 Pro reasoning 1,146
Muse Spark 1.1 minimal reasoning 1,143
Nemotron 3 Super low reasoning 1,141
MiniMax-M3 high reasoning 1,138
Qwen3.8 Max (0902) high reasoning 1,136
Grok 4.7 xhigh reasoning 1,133
Inkling 1,131
Step 5 Preview high reasoning 1,131
LongCat-2.0 medium reasoning 1,130
Mistral Small 3 1,129
Inkling-Small minimal reasoning 1,128
Step 3.5 Flash 1,124
Nemotron 3 Super medium reasoning 1,118
Solar Pro 3 medium reasoning 1,116
Qwen3.8 Max (0902) xhigh reasoning 1,115
gpt-oss-120b high reasoning 1,113
Granite 4.2 8B high reasoning 1,104
Mercury 2 low reasoning 1,102
Nemotron 3 Ultra low reasoning 1,087
Claude Sonnet 5.5 xhigh reasoning 1,086
Kimi K3 medium reasoning 1,082
Muse Spark 1.3 high reasoning 1,071
Mercury 2 1,057
MiMo-V2.5-Pro high reasoning 1,049
Claude Opus 5.5 max reasoning 1,045
Mistral Medium 3.5 medium reasoning 1,027
GPT-6.1 Sol max reasoning 1,000
Solar Pro 3 low reasoning 988
Nemotron 3 Nano Omni 30B-A3B reasoning 983

7 models have held number one on the official leaderboard since 2023

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

Perplexity Mistral AI xAI Xiaomi Corp NVIDIA MiniMax Meituan
Own leaderboard 7 leaders
LongCat-2.0 1,375

Source: Official StudyArena Text leaderboard

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
900 1,000 1,100 1,200 1,300 1,400 2025 2026 900 1,000 1,100 1,200 1,300 1,400 2025 2026

Source: Official StudyArena Text leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Instructions and chat

All 81 in Instructions and chat
  • IFEval In use Models answer about 500 prompts containing verifiable instructions such as word limits or required keywords. Code checks each instruction for strict and loose accuracy. Gemma 4 31B IT 98.9%
  • IFBench In use 58 new out-of-domain verifiable output constraints. Scored by programmatic constraint checks (strict or loose). Nemotron-Cascade 2 30B-A3B 82.9%
  • MultiChallenge In use Multi-turn conversations testing instruction retention, inference memory, versioned editing and self-coherence. Graded by LLM judge with rubrics. GPT-6.1 Sol 82%
  • Multi-IF In use Models follow verifiable instructions over three-turn conversations in eight languages, extending IFEval. Scored by instruction accuracy at each turn. LFM3-1.7B 89.1%
  • AdvancedIF In use Models respond to over 1,600 prompts with complex, multi-turn, and system-level instructions. A judge model checks each response against expert-written rubrics. MAI-Thinking-1 85%
  • IFStruct In use Models generate JSON or YAML that must follow a requested schema and phrasing constraints, without constrained decoding. A response passes only if every structural constraint holds. Gemma 4 31B IT 95.9%