Artificial Analysis

Artificial Analysis Long Context Reasoning

An Artificial Analysis long-context reasoning evaluation.

In index
Reasoning
Source ↗

Models218

Top result88.7%Kimi K3

Top-3 spread2.4 pts

Year2026

Result and price

143 of 218 results

Results

218 results

#ModelResult

01Kimi K388.7%

02Step 5 Preview88.3%

03Kimi K3.1 Pro86.3%

03MiMo-V2.6-Pro86.3%

05Claude Fable 5.185.3%

06Claude Opus 5.584.7%

07GPT-5.584.3%

08GPT-5.6 Sol84.0%

08DeepSeek V4.1 Flash 101784.0%

10Muse Wonder Full83.7%

10GPT-6 Sol83.7%

10GPT-5.6 Luna83.7%

13Claude Fable 5.283.3%

13Inkling 2 Small83.3%

13GPT-6 Luna83.3%

13GPT-5.3 Codex83.3%

13Muse Glimmer 30B83.3%

18GPT-6.1 Sol83.0%

18Muse Spark 1.383.0%

18GPT-5.6 Terra83.0%

18Ling 3.1 Flash83.0%

18MiniMax M383.0%

23Claude Sonnet 5.582.7%

23GPT-6.1 Terra82.7%

23Claude Haiku 5.582.7%

23GPT-5.282.7%

27Claude Fable 582.3%

27GPT-5.2-Codex82.3%

29GPT-5.482.0%

29Claude Sonnet 582.0%

29Gemini 3.1 Pro82.0%

29Qwen3.8-27B82.0%

33Gemini 3.7 Flash81.7%

34Gemini 3.8 Flash81.3%

35Sage-A1.581.0%

35Kimi K2.681.0%

37GPT-6 Astra80.7%

37Muse Wonder80.7%

37Qwen 3.6 Max (preview)80.7%

40Grok 4.680.3%

40DeepSeek V4 Pro 081380.3%

40Qwen3.8 Max Preview80.3%

43Gemini 3.6 Flash80.0%

43GPT-5.180.0%

45Gemini 4 Argon79.7%

45MiniMax M3.1 Flash79.7%

45GLM-5.379.7%

45Qwen3.8-Flash-Next79.7%

45DeepSeek V4 Flash 073179.7%

45MiMo-V2.5-Pro79.7%

51Beam 1.0 501B-A23B FP879.4%

52Gemini 4 Carbon79.3%

52Claude Opus 579.3%

52Grok 4.579.3%

52Apodex 1.179.3%

52Kimi K2.7 Code79.3%

52Apodex 1.1 Mini79.3%

58Muse Spark 1.279.0%

58Qwen3.7 Max79.0%

58Hy379.0%

61Claude Opus 4.7 (Adaptive)78.7%

61Beam 1.0 501B-A23B NVFP478.7%

63Muse Spark Open78.3%

63GLM-5.278.3%

63Qwen3.6 Plus78.3%

63MiniMax M2.778.3%

63Ling 3.0 Flash VL78.3%

68Beam 1.0 501B-A23B78.2%

68GPT-5 (high)78.2%

70Claude Opus 4.6 (Adaptive)78.0%

70Muse Spark78.0%

70Kimi K2.5 (Reasoning)78.0%

70Kimi K2.578.0%

74Claude Opus 4.877.7%

74Muse Spark 1.177.7%

74Qwen3.5-27B77.7%

77Claude Opus 4.5 Thinking77.3%

77Qwen3.5 397B (Reasoning)77.3%

77Inkling77.3%

77Qwen3.5 397B77.3%

77Qwen3.6-27B77.3%

82GPT-5.4 mini77.0%

82Ternary Bonsai 2 27B77.0%

84Grok 4.776.7%

84GPT-5.4 nano76.7%

86Qwen3.5-122B-A10B76.3%

86Quasar 438B76.3%

88Gemini 3 Pro76.0%

88GPT-5 (medium)76.0%

88Gemini 3.5 Flash-Lite76.0%

88Claude 4.1 Opus Thinking76.0%

92Claude Opus 4.775.7%

92Inkling-Small75.7%

92GLM-575.7%

95Ripple-65B Preview75.0%

95MiMo-V2-Omni75.0%

97o374.7%

98MiMo-V2.6-Flash74.3%

99Grok 4.1 Fast (Reasoning)74.0%

100GLM-5.173.7%

100Step 3.7 Flash73.7%

100Grok 4 Fast (Reasoning)73.7%

103Qwen3.7 Plus73.0%

103Ling 3.0 Flash73.0%

105Qwen3.5-35B-A3B72.0%

106GLM-5-Turbo71.7%

106Qwen3.6-35B-A3B71.7%

108Solar Pro 471.0%

108GLM-4.771.0%

110Claude Opus 4.570.7%

111Ling 3.0 Flash Sante70.5%

112GLM-5V-Turbo70.3%

113Gemma 4 31B69.7%

113Nemotron 3.5 Bolt 74B-A9B69.7%

115Gemini 3.5 Flash69.3%

115GPT-5.1-Codex-Max69.3%

115GPT-5.1-Codex69.3%

115Mistral Medium 3.5 128B69.3%

115Claude 4 Opus Thinking69.3%

120Gemini 2.5 Pro69.0%

121Nemotron 3.5 Super VL 120B A12B68.6%

122Claude Sonnet 4.668.3%

122MiMo-V2-Pro68.3%

122GPT-4.168.3%

125Grok 468.0%

125Mercury 2.568.0%

127Nemotron 3.5 Super VL 120B A12B NVFP467.7%

128Nemotron 3.5 Bolt 74B-A9B BF1667.4%

129Spark-X2.6-4B67.3%

130Claude Opus 4.667.0%

130Nemotron 3 Ultra67.0%

132Hy3 Preview66.7%

133A.X K266.0%

134Gemma 4 26B A4B65.7%

135o165.0%

136Grok 4.364.3%

137Gemma 4 12B63.7%

138Qwen3.5-4B63.0%

139Solar Open 262.3%

140K-Exaone61.3%

141o4-mini61.0%

142Ling 3.0 Tiny60.3%

143MiniCPM5-2B59.0%

144DeepSeek V3.1 (Reasoning)56.7%

145K-EXAONE 2.056.2%

146DeepSeek-R155.7%

147Gemini 3 Flash55.3%

148Grok Code Fast 153.0%

148Kimi K253.0%

150LFM3-1.7B52.8%

151Command A+52.7%

152GPT-OSS 120B52.0%

153Claude 3.7 Sonnet51.7%

154Qwen3 Max50.0%

154Llama 4 Maverick50.0%

156Gemini 2.5 Flash49.9%

157Mistral Small 449.7%

—Mistral Small 4 (Reasoning)49.7%

158GPT-4o49.3%

159Nemotron 3.5 Lightning 30B A3B NVFP449.2%

160Granite 4.2 30B49.0%

161DeepSeek V3.147.0%

162GLM-4.5-Air46.7%

163DeepSeek V3.245.7%

164Granite 4.2 8B45.0%

165Claude 4 Sonnet44.0%

165GPT-4.1 mini44.0%

167DeepSeek V3 032440.7%

168Nemotron 3 Nano Omni 30B A3B39.7%

169Trinity-Large-Thinking38.0%

169Trinity-Large-Preview38.0%

169Celeris-138.0%

172North Mini Code37.3%

173MiMo-V2-Flash36.3%

174Mistral Large 336.0%

175GPT-OSS 20B34.7%

176Solar Pro 332.3%

177Gemma 4 E4B32.0%

177Gemini 2.5 Flash-Lite32.0%

179Ling 2.6 Flash31.3%

179Grok 4.1 Fast31.3%

179Gemini 2.0 Flash31.3%

179Mistral Medium 331.3%

183DeepSeek V329.3%

184Llama 4 Scout27.7%

184Claude 3 Haiku27.7%

186Claude 3.5 Haiku27.3%

187QwQ-32B26.7%

188GLM-4.626.3%

189Llama 3.1 405B25.3%

190Granite 4.2 3B24.3%

191Mistral Small 3.122.3%

192Nova Pro21.0%

193GPT-4.1 nano20.3%

194Devstral Small18.7%

195Gemma 4 E2B16.3%

196Llama 3.3 70B15.7%

196Ultravox v0.6 Llama 3.3 70B15.7%

198Nemotron 3 Nano 30B10.7%

199DeepSeek R1 Distill Qwen 32B8.7%

200Gemma 3 27B7.3%

201Granite-4.0-H-1B7.0%

202LFM2.5-2.6B5.7%

203Llama 3.1 70B4.7%

204Sarvam 105B0.0%

204Magistral Medium0.0%

204Magistral Small0.0%

204LFM2.5-8B-A1B0.0%

204Sarvam 30B0.0%

204Phi-4-reasoning0.0%

204Phi-40.0%

204Granite-4.0-H-350M0.0%

204Granite-4.0-350M0.0%

204Exaone 4.0 1.2B0.0%

204LFM2.5-VL-1.6B-Extract0.0%

204Qwen3-Omni-30B-A3B-Instruct0.0%

204Qwen3-Omni-30B-A3B-Thinking0.0%

204Solar Pro 20.0%