Inference

Artificial Analysis Long Context Reasoning

An Artificial Analysis long-context reasoning evaluation.

  • In index
  • Reasoning
ModelsModels with a result.177
Top resultBest result on this test.88.7%Kimi K3
Top-3 spreadPoints from first to third.2.4 pts
YearYear of release.N/A

Result and price

020406080100$0.03$0.1$0.3$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS · LOG SCALE

Results

177 results
#ModelResultThe score from the source.
01Kimi K388.7%
02Step 5 Preview88.3%
03MiMo-V2.6-Pro86.3%
04Claude Fable 5.185.3%
05Claude Opus 5.584.7%
06GPT-5.584.3%
07GPT-5.6 Sol84.0%
07DeepSeek V4.1 Flash84.0%
09GPT-6 Sol83.7%
09GPT-5.6 Luna83.7%
11GPT-6 Luna83.3%
11GPT-5.3 Codex83.3%
11Muse Glimmer 30B83.3%
14Muse Spark 1.383.0%
14GPT-5.6 Terra83.0%
14MiniMax M383.0%
17GPT-5.282.7%
18Claude Fable 582.3%
18GPT-5.2-Codex82.3%
20GPT-5.482.0%
20Claude Sonnet 582.0%
20Gemini 3.1 Pro82.0%
20Qwen3.8-27B82.0%
24Gemini 3.7 Flash81.7%
25Gemini 3.8 Flash81.3%
26Kimi K2.681.0%
27GPT-6 Astra80.7%
27Qwen 3.6 Max (preview)80.7%
29Grok 4.680.3%
29DeepSeek V4 Pro 081380.3%
29Qwen3.8 Max Preview80.3%
32Gemini 3.6 Flash80.0%
32GPT-5.180.0%
34GLM-5.379.7%
34Qwen3.8-Flash-Next79.7%
34DeepSeek V4 Flash 073179.7%
34MiMo-V2.5-Pro79.7%
38Claude Opus 579.3%
38Grok 4.579.3%
38Apodex 1.179.3%
38Kimi K2.7 Code79.3%
38Apodex 1.1 Mini79.3%
43Muse Spark 1.279.0%
43Qwen3.7 Max79.0%
43Hy379.0%
46Claude Opus 4.7 (Adaptive)78.7%
47GLM-5.278.3%
47Qwen3.6 Plus78.3%
47MiniMax M2.778.3%
47Ling 3.0 Flash VL78.3%
51GPT-5 (high)78.2%
52Claude Opus 4.6 (Adaptive)78.0%
52Muse Spark78.0%
52Kimi K2.5 (Reasoning)78.0%
52Kimi K2.578.0%
56Claude Opus 4.877.7%
56Muse Spark 1.177.7%
56Qwen3.5-27B77.7%
59Claude Opus 4.5 Thinking77.3%
59Inkling77.3%
59Qwen3.6-27B77.3%
62GPT-5.4 mini77.0%
62Ternary Bonsai 2 27B77.0%
64Grok 4.776.7%
64GPT-5.4 nano76.7%
66Quasar 438B76.3%
66Qwen3.5-122B-A10B76.3%
68Gemini 3 Pro76.0%
68GPT-5 (medium)76.0%
68Gemini 3.5 Flash-Lite76.0%
68Claude 4.1 Opus Thinking76.0%
72Claude Opus 4.775.7%
72Inkling-Small75.7%
72GLM-575.7%
75MiMo-V2-Omni75.0%
76o374.7%
77Grok 4.1 Fast (Reasoning)74.0%
78GLM-5.173.7%
78Step 3.7 Flash73.7%
78Grok 4 Fast (Reasoning)73.7%
81Qwen3.7 Plus73.0%
81Ling 3.0 Flash73.0%
81Ling 3.0 Flash FP873.0%
84Qwen3.5-35B-A3B72.0%
85GLM-5-Turbo71.7%
85Qwen3.6-35B-A3B71.7%
87Solar Pro 471.0%
87GLM-4.771.0%
89Claude Opus 4.570.7%
90GLM-5V-Turbo70.3%
91Gemma 4 31B69.7%
92Gemini 3.5 Flash69.3%
92GPT-5.1-Codex-Max69.3%
92GPT-5.1-Codex69.3%
92Mistral Medium 3.5 128B69.3%
96Gemini 2.5 Pro69.0%
97Claude Sonnet 4.668.3%
97MiMo-V2-Pro68.3%
97GPT-4.168.3%
100Grok 468.0%
100Mercury 2.568.0%
102Claude Opus 4.667.0%
102Nemotron 3 Ultra67.0%
104Hy3 Preview66.7%
105A.X K266.0%
106Gemma 4 26B A4B65.7%
106Nemotron 3 Super 100B65.7%
108o165.0%
109Grok 4.364.3%
109Qwen3.5 397B64.3%
109Qwen3.5 397B (Reasoning)64.3%
112Gemma 4 12B63.7%
113Solar Open 262.3%
114K-Exaone61.3%
115Ling 3.0 Tiny60.3%
116MiniCPM5-2B59.0%
117DeepSeek V3.1 (Reasoning)56.7%
118K-EXAONE 2.056.2%
119DeepSeek-R155.7%
120Gemini 3 Flash55.3%
121Kimi K253.0%
121Grok Code Fast 153.0%
123Command A+52.7%
124GPT-OSS 120B52.0%
125Qwen3 Max50.0%
125Llama 4 Maverick50.0%
127Gemini 2.5 Flash49.9%
128Mistral Small 449.7%
—Mistral Small 4 (Reasoning)49.7%
129GPT-4o49.3%
130Nemotron 3.5 Lightning 30B A3B NVFP449.2%
131Granite 4.2 30B49.0%
132DeepSeek V3.147.0%
133GLM-4.5-Air46.7%
134DeepSeek V3.245.7%
135Granite 4.2 8B45.0%
136Claude 4 Sonnet44.0%
136GPT-4.1 mini44.0%
138Nemotron 3 Nano Omni 30B A3B39.7%
139Trinity-Large-Thinking38.0%
139Trinity-Large-Preview38.0%
139Nemotron 3 Nano 30B38.0%
139Celeris-138.0%
143North Mini Code37.3%
144MiMo-V2-Flash36.3%
145Mistral Large 336.0%
146GPT-OSS 20B34.7%
147Solar Pro 332.3%
148Gemma 4 E4B32.0%
149Ling 2.6 Flash31.3%
149Grok 4.1 Fast31.3%
149Mistral Medium 331.3%
152DeepSeek V329.3%
153Llama 4 Scout27.7%
153Claude 3 Haiku27.7%
155GLM-4.626.3%
156Llama 3.1 405B25.3%
157Granite 4.2 3B24.3%
158Nova Pro21.0%
159GPT-4.1 nano20.3%
160Gemma 4 E2B16.3%
161Ultravox v0.6 Llama 3.3 70B15.7%
162DeepSeek R1 Distill Qwen 32B8.7%
163Gemma 3 27B7.3%
164Granite-4.0-H-1B7.0%
165LFM2.5-2.6B5.7%
166LFM2.5-8B-A1B0.0%
166Qwen3-Omni-30B-A3B-Thinking0.0%
166Qwen3-Omni-30B-A3B-Instruct0.0%
166Phi-40.0%
166LFM2.5-VL-1.6B-Extract0.0%
166Exaone 4.0 1.2B0.0%
166Granite-4.0-350M0.0%
166Granite-4.0-H-350M0.0%
166Sarvam 105B0.0%
166Sarvam 30B0.0%
166Solar Pro 20.0%