Long context reasoning (AA-LCR) leaderboard
- Kimi K3 leads
- 88.7%
- models measured
- 178
- lead over #2
- +0.3 pts
- median
- 73.0%
Leader on Long context reasoning (AA-LCR) · read October 21
On October 21, 2026, 179 of our catalog's models had a Long context reasoning (AA-LCR) result, and MoonshotAI Kimi K3 led with 88.7%, ahead of StepFun Step 5 Preview at 88.3%.
Questions that need reasoning across long documents, around 100K tokens. Share answered correctly. Run by Artificial Analysis on its own harness.
Role in the index
Long context reasoning (AA-LCR) is shown as published and does not feed any of our indexes. Read the methodology
| # | Model | Result | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | Kimi K3Moonshot AI | 88.7% | 62.3 | $15.00 | |
| 2 | Step 5 PreviewStepFun | 88.3% | 61.1 | $2.70 | |
| 3 | MiMo-V2.6-ProXiaomi | 86.3% | 65.4 | $0.87 | |
| 4 | Claude Fable 5.1Anthropic | 85.3% | 84.2 | $50.00 | |
| 5 | Claude Opus 5.5Anthropic | 84.7% | 88.8 | $20.00 | |
| 6 | GPT-5.5OpenAI | 84.3% | 71.3 | $30.00 | |
| 7 | DeepSeek V4.1 FlashDeepSeek | 84.0% | 62.5 | $0.78 | |
| 8 | Gemini 3.8 FlashGoogle | 84.0% | 61.6 | $3.75 | |
| 9 | GPT-5.6 SolOpenAI | 84.0% | 73.5 | $20.00 | |
| 10 | GPT-6.1 SolOpenAI | 84.0% | 82.4 | $10.00 | |
| 11 | GPT-5.6 LunaOpenAI | 83.7% | 58.0 | $1.20 | |
| 12 | GPT-6 SolOpenAI | 83.7% | 76.0 | $10.00 | |
| 13 | GPT-5.3-CodexOpenAI | 83.3% | 60.0 | $14.00 | |
| 14 | GPT-6 LunaOpenAI | 83.3% | 57.3 | $0.50 | |
| 15 | Solar Mini 4Upstage | 83.3% | Not benchmarked | $0.20 | |
| 16 | Gemini 3.7 FlashGoogle | 83.0% | 58.0 | $3.75 | |
| 17 | Ling 3.1 FlashinclusionAI | 83.0% | Not benchmarked | $0.225 | |
| 18 | Muse Spark 1.3Meta | 83.0% | 57.4 | $4.25 | |
| 19 | MiniMax M3MiniMax | 83.0% | 39.2 | $1.20 | |
| 20 | GPT-5.6 TerraOpenAI | 83.0% | 63.9 | $12.00 | |
| 21 | Claude Haiku 5.5Anthropic | 82.7% | 70.7 | $0.50 | |
| 22 | Claude Sonnet 5.5Anthropic | 82.7% | 87.3 | $10.00 | |
| 23 | DeepSeek V4.1 Flash 1017DeepSeek | 82.7% | 70.3 | $0.60 | |
| 24 | GPT-5.2OpenAI | 82.7% | 55.9 | $14.00 | |
| 25 | Claude Fable 5Anthropic | 82.3% | 79.3 | $50.00 | |
| 26 | GPT-5.2-CodexOpenAI | 82.3% | 45.4 | $14.00 | |
| 27 | Claude Opus 5Anthropic | 82.0% | 83.0 | $25.00 | |
| 28 | Claude Sonnet 5Anthropic | 82.0% | 57.7 | $10.00 | |
| 29 | Gemini 3.1 Pro PreviewGoogle | 82.0% | 32.3 | $12.00 | |
| 30 | GPT-5.4OpenAI | 82.0% | 64.4 | $15.00 | |
| 31 | Qwen3.8 27BQwen | 82.0% | 74.2 | $2.55 | |
| 32 | DeepSeek V4 Flash Vision ExpDeepSeek | 81.3% | Not benchmarked | $1.32 | |
| 33 | Mistral Large 4Mistral AI | 81.3% | 56.4 | $4.18 | |
| 34 | Kimi K2.6Moonshot AI | 81.0% | 52.1 | $4.00 | |
| 35 | Grok 4.6xAI | 81.0% | 65.7 | $6.00 | |
| 36 | Muse WonderMeta | 80.7% | 71.9 | $20.00 | |
| 37 | GPT-6 AstraOpenAI | 80.7% | 85.1 | $50.00 | |
| 38 | DeepSeek V4 Pro 0813DeepSeek | 80.3% | 56.1 | $1.98 | |
| 39 | Qwen3.8 2.4T A95BQwen | 80.3% | Not benchmarked | $6.00 | |
| 40 | Qwen3.8 Max (0902)Qwen | 80.3% | 62.4 | $6.00 | |
| 41 | Claude Sonnet 4.6Anthropic | 80.0% | 57.3 | $15.00 | |
| 42 | Gemini 3.6 FlashGoogle | 80.0% | 62.5 | $3.75 | |
| 43 | GPT-5.1OpenAI | 80.0% | 40.4 | $10.00 | |
| 44 | Grok 4.7xAI | 80.0% | 66.7 | $6.00 | |
| 45 | GLM 5.3 FlashZ.ai | 80.0% | 57.7 | $0.50 | |
| 46 | DeepSeek V4 Flash 0731DeepSeek | 79.7% | 51.8 | $0.28 | |
| 47 | Gemini 4 ArgonGoogle | 79.7% | 81.0 | $10.00 | |
| 48 | MiMo-V2.5-ProXiaomi | 79.7% | 31.6 | $0.9152 | |
| 49 | GLM 5.3Z.ai | 79.7% | 70.6 | $4.40 | |
| 50 | Kimi K2.7 CodeMoonshot AI | 79.3% | 53.4 | $4.00 | |
| 51 | Nemotron 3 UltraNVIDIA | 79.3% | 30.0 | $2.40 | |
| 52 | Grok 4.5xAI | 79.3% | 58.6 | $6.00 | |
| 53 | Muse Spark 1.2Meta | 79.0% | 55.7 | $4.25 | |
| 54 | Qwen3.7 MaxQwen | 79.0% | 54.1 | $4.425 | |
| 55 | Hy3Tencent | 79.0% | Not benchmarked | $0.33 | |
| 56 | Claude Opus 4.7Anthropic | 78.7% | 67.5 | $25.00 | |
| 57 | Ling 3.0 Flash VLinclusionAI | 78.3% | Not benchmarked | $0.18 | |
| 58 | Muse Spark OpenMeta | 78.3% | 47.4 | $4.25 | |
| 59 | MiniMax M2.7MiniMax | 78.3% | 36.2 | $1.20 | |
| 60 | Qwen3.6 PlusQwen | 78.3% | 44.1 | $1.95 | |
| 61 | GLM 5.2Z.ai | 78.3% | 61.1 | $4.40 | |
| 62 | GPT-5OpenAI | 78.2% | 38.8 | $10.00 | |
| 63 | Claude Opus 4.6Anthropic | 78.0% | 57.2 | $25.00 | |
| 64 | Gemini 3 Flash PreviewGoogle | 78.0% | 43.7 | $3.00 | |
| 65 | Kimi K2.5Moonshot AI | 78.0% | 38.6 | $2.85 | |
| 66 | Claude Opus 4.8Anthropic | 77.7% | 65.0 | $25.00 | |
| 67 | Muse Spark 1.1Meta | 77.7% | 70.3 | $4.25 | |
| 68 | Qwen3.5-27BQwen | 77.7% | Not benchmarked | $1.56 | |
| 69 | Claude Opus 4.5Anthropic | 77.3% | 44.0 | $25.00 | |
| 70 | Qwen3.5 397B A17BQwen | 77.3% | Not benchmarked | $2.34 | |
| 71 | Qwen3.6 27BQwen | 77.3% | 36.8 | $2.70 | |
| 72 | T InklingThinking Machines | 77.3% | 31.5 | $4.05 | |
| 73 | GPT-5.4 MiniOpenAI | 77.0% | 35.8 | $4.50 | |
| 74 | GPT-5.4 NanoOpenAI | 76.7% | 44.8 | $1.25 | |
| 75 | Qwen3.5-122B-A10BQwen | 76.3% | Not benchmarked | $2.08 | |
| 76 | Gemini 3.5 Flash LiteGoogle | 76.0% | 49.0 | $2.50 | |
| 77 | T Inkling SmallThinking Machines | 75.7% | 27.3 | $1.20 | |
| 78 | GLM 5Z.ai | 75.7% | 41.7 | $3.20 | |
| 79 | Grok 4.3xAI | 75.0% | 37.4 | $2.50 | |
| 80 | o3OpenAI | 74.7% | Not benchmarked | $8.00 | |
| 81 | Gemini 3.1 Flash LiteGoogle | 74.3% | Not benchmarked | $1.50 | |
| 82 | Gemini 3.5 FlashGoogle | 74.3% | 39.8 | $9.00 | |
| 83 | MiMo-V2.6-FlashXiaomi | 74.3% | 54.7 | $0.28 | |
| 84 | Step 3.7 FlashStepFun | 73.7% | Not benchmarked | $1.15 | |
| 85 | GLM 5.1Z.ai | 73.7% | 49.2 | $4.40 | |
| 86 | DeepSeek V3.2DeepSeek | 73.3% | 30.0 | $0.88 | |
| 87 | MiniMax M2.5MiniMax | 73.3% | 37.2 | $1.20 | |
| 88 | Ling 3.0 FlashinclusionAI | 73.0% | 29.6 | $0.1215 | |
| 89 | Qwen3.7 PlusQwen | 73.0% | 33.2 | $1.28 | |
| 90 | MiMo-V2.5Xiaomi | 73.0% | 50.0 | $0.336 | |
| 91 | DeepSeek V3.2 ExpDeepSeek | 72.3% | Not benchmarked | $0.41 | |
| 92 | GPT-5 MiniOpenAI | 72.3% | 34.2 | $2.00 | |
| 93 | Kimi K2 ThinkingMoonshot AI | 72.0% | Not benchmarked | $2.50 | |
| 94 | Qwen3 235B A22B Instruct 2507Qwen | 72.0% | Not benchmarked | $0.75 | |
| 95 | Qwen3.5-35B-A3BQwen | 72.0% | Not benchmarked | $1.30 | |
| 96 | Mercury 2.5Inception | 71.7% | 1.6 | $0.15 | |
| 97 | Qwen3.6 35B A3BQwen | 71.7% | Not benchmarked | $1.00 | |
| 98 | GLM 5 TurboZ.ai | 71.7% | Not benchmarked | $4.00 | |
| 99 | GLM 4.7Z.ai | 71.0% | 55.1 | $2.20 | |
| 100 | Solar Pro 4Upstage | 70.7% | Not benchmarked | $0.60 | |
| 101 | GLM 5V TurboZ.ai | 70.3% | Not benchmarked | $4.00 | |
| 102 | Qwen3.5-9BQwen | 70.0% | Not benchmarked | $0.15 | |
| 103 | Gemma 4 31BGoogle | 69.7% | 39.0 | $0.40 | |
| 104 | DeepSeek V3.1 TerminusDeepSeek | 69.3% | Not benchmarked | $1.0142 | |
| 105 | Mistral Medium 3.5Mistral AI | 69.3% | 25.5 | $7.50 | |
| 106 | GPT-5.1-CodexOpenAI | 69.3% | 15.3 | $10.00 | |
| 107 | Gemini 2.5 ProGoogle | 69.0% | 19.1 | $10.00 | |
| 108 | Nemotron 3.5 Super VLNVIDIA | 69.0% | Not benchmarked | $0.50 | |
| 109 | Grok 4.20xAI | 69.0% | 33.3 | $2.50 | |
| 110 | GPT-4.1OpenAI | 68.3% | Not benchmarked | $8.00 | |
| 111 | MiniMax M2.1MiniMax | 67.7% | Not benchmarked | $1.20 | |
| 112 | GPT-5.1-Codex-MiniOpenAI | 66.7% | Not benchmarked | $2.00 | |
| 113 | Gemma 4 26B A4BGoogle | 65.7% | Not benchmarked | $0.60 | |
| 114 | Nemotron 3 SuperNVIDIA | 65.7% | Not benchmarked | $0.45 | |
| 115 | Gemini 2.5 FlashGoogle | 65.3% | Not benchmarked | $2.50 | |
| 116 | LongCat 2.0Meituan | 65.0% | Not benchmarked | $1.20 | |
| 117 | o1OpenAI | 65.0% | Not benchmarked | $60.00 | |
| 118 | Hy3 previewTencent | 64.7% | Not benchmarked | $0.60 | |
| 119 | MiniMax M2MiniMax | 64.3% | Not benchmarked | $1.02 | |
| 120 | Qwen3 Next 80B A3B InstructQwen | 63.7% | Not benchmarked | $1.10 | |
| 121 | Qwen3 VL 235B A22B InstructQwen | 63.7% | Not benchmarked | $1.90 | |
| 122 | Step 3.5 FlashStepFun | 63.0% | Not benchmarked | $0.30 | |
| 123 | Qwen3 30B A3B Instruct 2507Qwen | 61.3% | Not benchmarked | $0.30 | |
| 124 | o4 MiniOpenAI | 61.0% | Not benchmarked | $4.40 | |
| 125 | Nemotron 3.5 LightningNVIDIA | 60.3% | 22.6 | $0.147 | |
| 126 | Grok 3RetiredxAI | 58.0% | Not benchmarked | $15.00 | |
| 127 | R1 0528DeepSeek | 55.7% | Not benchmarked | $2.18 | |
| 128 | Gemini 2.5 Flash LiteGoogle | 55.7% | Not benchmarked | $0.40 | |
| 129 | GLM 4.6Z.ai | 54.0% | 8.8 | $2.20 | |
| 130 | Kimi K2 0711Moonshot AI | 53.0% | Not benchmarked | $2.30 | |
| 131 | Kimi K2 0905Moonshot AI | 53.0% | Not benchmarked | $2.50 | |
| 132 | Command ACohere | 52.7% | Not benchmarked | $10.00 | |
| 133 | GLM 4.5Z.ai | 52.7% | Not benchmarked | $2.20 | |
| 134 | gpt-oss-120bOpenAI | 52.0% | Not benchmarked | $0.50 | |
| 135 | Llama 4 MaverickMeta Llama | 50.0% | Not benchmarked | $1.00 | |
| 136 | Mistral Small 4Mistral AI | 49.7% | Not benchmarked | $0.60 | |
| 137 | GPT-4oOpenAI | 49.3% | Not benchmarked | $10.00 | |
| 138 | GLM 4.6VZ.ai | 48.7% | Not benchmarked | $0.90 | |
| 139 | Qwen3 Coder NextQwen | 47.0% | Not benchmarked | $0.80 | |
| 140 | GLM 4.5 AirZ.ai | 46.7% | Not benchmarked | $1.10 | |
| 141 | Qwen3 Coder 480B A35BQwen | 45.7% | Not benchmarked | $1.50 | |
| 142 | Granite 4.2 8BIBM Granite | 45.0% | Not benchmarked | $0.20 | |
| 143 | GPT-5 NanoOpenAI | 45.0% | Not benchmarked | $0.40 | |
| 144 | GPT-4.1 MiniOpenAI | 44.0% | Not benchmarked | $1.60 | |
| 145 | Mercury 2Inception | 43.7% | Not benchmarked | $0.75 | |
| 146 | o3 MiniOpenAI | 43.0% | Not benchmarked | $4.40 | |
| 147 | Mistral Medium 3.1Mistral AI | 42.7% | Not benchmarked | $2.00 | |
| 148 | GLM 4.7 FlashZ.ai | 41.7% | Not benchmarked | $0.40 | |
| 149 | GPT-4o (2024-08-06)OpenAI | 41.0% | Not benchmarked | $10.00 | |
| 150 | DeepSeek V3 0324DeepSeek | 40.7% | Not benchmarked | $1.07 | |
| 151 | Trinity Large ThinkingArcee AI | 38.0% | Not benchmarked | $0.80 | |
| 152 | Nemotron 3 Nano 30B A3BNVIDIA | 38.0% | Not benchmarked | $0.22 | |
| 153 | gpt-oss-20bOpenAI | 34.7% | Not benchmarked | $0.15 | |
| 154 | Nova Premier 1.0Amazon | 34.3% | Not benchmarked | $12.50 | |
| 155 | Qwen3 VL 8B InstructQwen | 33.3% | Not benchmarked | $0.75 | |
| 156 | Qwen3 Coder 30B A3B InstructQwen | 32.7% | Not benchmarked | $0.44 | |
| 157 | Solar Pro 3Upstage | 32.3% | Not benchmarked | $0.60 | |
| 158 | Mistral Medium 3Mistral AI | 31.3% | Not benchmarked | $2.00 | |
| 159 | Llama 4 ScoutMeta Llama | 27.7% | Not benchmarked | $0.50 | |
| 160 | Nova Pro 1.0Amazon | 21.0% | Not benchmarked | $3.20 | |
| 161 | GPT-4.1 NanoOpenAI | 20.3% | Not benchmarked | $0.40 | |
| 162 | Nova Lite 1.0Amazon | 19.0% | Not benchmarked | $0.24 | |
| 163 | Llama 3.1 8B InstructMeta Llama | 18.0% | Not benchmarked | $0.08 | |
| 164 | Llama 3.3 70B InstructMeta Llama | 15.7% | Not benchmarked | $0.72 | |
| 165 | Nova Micro 1.0Amazon | 13.0% | Not benchmarked | $0.14 | |
| 166 | Gemma 3 12BGoogle | 8.3% | Not benchmarked | $0.1625 | |
| 167 | Gemma 3 27BGoogle | 7.3% | Not benchmarked | $0.30 | |
| 168 | Granite 4.0 MicroIBM Granite | 7.0% | Not benchmarked | $0.112 | |
| 169 | Gemma 3 4BGoogle | 6.7% | Not benchmarked | $0.075 | |
| 170 | Llama 3.2 1B InstructMeta Llama | 6.7% | Not benchmarked | $0.201 | |
| 171 | Llama 3.2 3B InstructMeta Llama | 4.3% | Not benchmarked | $0.3325 | |
| 172 | Mistral Large 2407Mistral AI | 2.0% | Not benchmarked | $6.00 | |
| 173 | Phi 4Microsoft | 0.0% | Not benchmarked | $0.14 | |
| 174 | Mistral Small 3Mistral AI | 0.0% | Not benchmarked | $0.08 | |
| 175 | Qwen3 14BQwen | 0.0% | Not benchmarked | $0.23 | |
| 176 | Qwen3 30B A3BQwen | 0.0% | Not benchmarked | $0.50 | |
| 177 | Qwen3 32BQwen | 0.0% | Not benchmarked | $0.40 | |
| 178 | Reka Flash 3Reka | 0.0% | Not benchmarked | $0.20 | |
| 179 | GLM 4.5VZ.ai | 0.0% | Not benchmarked | $1.80 |
Only models in our catalog are listed, and each appears once with its best published run. 198 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.