# Artificial Analysis Humanity's Last Exam (AA-HLE)

> A display-only Artificial Analysis Humanity's Last Exam score.

Canonical page: https://benchlm.ai/benchmarks/aahle

- Category: [Knowledge](/knowledge)
- Last updated: September 10, 2026

## About AA-HLE

- Year: 2026
- Tasks: Expert-level questions
- Format: Accuracy
- Difficulty: Frontier expert reasoning
- Paper: [Artificial Analysis Humanity's Last Exam Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/hle)

BenchLM stores the Artificial Analysis HLE result separately from the weighted HLE lane so AA refreshes remain display-only.

AA-HLE is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (181 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [Claude Fable 5.1](/models/claude-fable-5-1) | Anthropic | 59.1% |
| 2 | [Claude Fable 5](/models/claude-fable) | Anthropic | 55.5% |
| 3 | [Claude Opus 5](/models/claude-opus-5) | Anthropic | 54.9% |
| 4 | [GPT-6 Astra](/models/gpt-6-astra) | OpenAI | 54.7% |
| 5 | [GPT-5.6 Sol](/models/gpt-5-6-sol) | OpenAI | 49.5% |
| 6 | [Claude Opus 4.8](/models/claude-opus-4-8) | Anthropic | 48.7% |
| 7 | [Muse Spark 1.3](/models/muse-spark-1-3) | Meta | 48.7% |
| 8 | [Gemini 3.7 Flash](/models/gemini-3-7-flash) | Google | 47.9% |
| 9 | [Gemini 3.8 Flash](/models/gemini-3-8-flash) | Google | 47.8% |
| 10 | [Gemini 3.1 Pro](/models/gemini-3-1-pro) | Google | 47.0% |
| 11 | [Kimi K3](/models/kimi-k3) | Moonshot AI | 46.9% |
| 12 | [Muse Spark 1.1](/models/muse-spark-1-1) | Meta | 46.2% |
| 13 | [GPT-5.5](/models/gpt-5-5) | OpenAI | 45.8% |
| 14 | [Muse Spark 1.2](/models/muse-spark-1-2) | Meta | 45.5% |
| 15 | [GPT-5.4](/models/gpt-5-4) | OpenAI | 43.7% |
| 16 | [Qwen3.8 Max Preview](/models/qwen3-8-max-preview) | Alibaba | 43.0% |
| 17 | [GPT-5.6 Terra](/models/gpt-5-6-terra) | OpenAI | 42.9% |
| 18 | [Grok 4.6](/models/grok-4-6) | xAI | 42.9% |
| 19 | [Gemini 3.5 Flash](/models/gemini-3-5-flash) | Google | 42.7% |
| 20 | [Grok 4.5](/models/grok-4-5) | xAI | 42.7% |
| 21 | [GPT-5.3 Codex](/models/gpt-5-3-codex) | OpenAI | 42.5% |
| 22 | [Claude Opus 4.7 (Adaptive)](/models/claude-opus-4-7-adaptive) | Anthropic | 42.3% |
| 23 | [GLM-5.3](/models/glm-5-3) | Z.AI | 42.3% |
| 24 | [Claude Sonnet 5](/models/claude-sonnet-5) | Anthropic | 41.3% |
| 25 | [GLM-5.2](/models/glm-5-2) | Z.AI | 41.1% |
| 26 | [DeepSeek V4 Pro 0813](/models/deepseek-v4-pro-0813) | DeepSeek | 41.0% |
| 27 | [Gemini 3.6 Flash](/models/gemini-3-6-flash) | Google | 40.8% |
| 28 | [Muse Spark](/models/muse-spark) | Meta | 40.7% |
| 29 | [Qwen3.7 Max](/models/qwen3-7-max) | Alibaba | 40.5% |
| 30 | [GLM-5.3-Flash](/models/glm-5-3-flash) | Z.AI | 39.9% |
| 31 | [Claude Opus 4.6 (Adaptive)](/models/claude-opus-4-6-thinking) | Anthropic | 39.9% |
| 32 | [Gemini 3 Pro](/models/gemini-3-pro) | Google | 39.7% |
| 33 | [GPT-5.6 Luna](/models/gpt-5-6-luna) | OpenAI | 39.5% |
| 34 | [DeepSeek V4.1 Flash](/models/deepseek-v4-1-flash) | DeepSeek | 39.2% |
| 35 | [MiniMax M3](/models/minimax-m3) | MiniMax | 39.0% |
| 36 | [DeepSeek V4 Flash 0731](/models/deepseek-v4-flash-0731) | DeepSeek | 38.6% |
| 37 | [Qwen3.8-Flash-Next](/models/qwen3-8-flash-next) | Alibaba | 38.0% |
| 38 | [GPT-5.2](/models/gpt-5-2) | OpenAI | 37.7% |
| 39 | [Kimi K2.6](/models/kimi-2-6) | Moonshot AI | 37.5% |
| 40 | [Grok 4.3](/models/grok-4-3) | xAI | 37.2% |
| 41 | [MiMo-V2.5-Pro](/models/mimo-v2-5-pro) | Xiaomi | 35.7% |
| 42 | [GPT-5.2-Codex](/models/gpt-5-2-codex) | OpenAI | 35.7% |
| 43 | [Qwen3.7 Plus](/models/qwen3-7-plus) | Alibaba | 35.6% |
| 44 | [Kimi K2.7 Code](/models/kimi-k2-7-code) | Moonshot AI | 35.0% |
| 45 | [Apodex 1.1](/models/apodex-1-1) | Apodex | 34.1% |
| 46 | [Apodex 1.1 Mini](/models/apodex-1-1-mini) | Apodex | 34.1% |
| 47 | [Qwen3.8-27B](/models/qwen3-8-27b) | Alibaba | 33.9% |
| 48 | [Hy3](/models/hy3) | Tencent | 33.5% |
| 49 | [Hy3 Preview](/models/hy3-preview) | Tencent | 33.5% |
| 50 | [Claude Opus 4.7](/models/claude-opus-4-7) | Anthropic | 33.3% |
| 51 | [Inkling-Small](/models/inkling-small) | Thinking Machines Lab | 33.3% |
| 52 | [Inkling](/models/inkling) | Thinking Machines Lab | 31.9% |
| 53 | [Qwen 3.6 Max (preview)](/models/qwen3-6-max-preview) | Alibaba | 30.8% |
| 54 | [Kimi K2.5 (Reasoning)](/models/kimi-k2-5-reasoning) | Moonshot AI | 30.7% |
| 55 | [Kimi K2.5](/models/kimi-k2-5) | Moonshot AI | 30.7% |
| 56 | [MiMo-V2-Pro](/models/mimo-v2-pro) | Xiaomi | 30.4% |
| 57 | [GLM-5.1](/models/glm-5-1) | Z.AI | 30.1% |
| 58 | [Claude Opus 4.5 Thinking](/models/claude-opus-4-5-thinking) | Anthropic | 30.1% |
| 59 | [MiniMax M2.7](/models/minimax-m2-7) | MiniMax | 29.6% |
| 60 | [A.X K2](/models/a-x-k2) | SK Telecom | 29.6% |
| 61 | [GLM-5](/models/glm-5) | Z.AI | 29.3% |
| 62 | [Solar Pro 4](/models/solar-pro-4) | Upstage | 29.2% |
| 63 | [GPT-5.1](/models/gpt-5-1) | OpenAI | 28.5% |
| 64 | [GPT-5 (high)](/models/gpt-5-high) | OpenAI | 28.5% |
| 65 | [Nemotron 3 Ultra](/models/nemotron-3-ultra) | NVIDIA | 28.4% |
| 66 | [GPT-5.4 nano](/models/gpt-5-4-nano) | OpenAI | 28.3% |
| 67 | [GPT-5.4 mini](/models/gpt-5-4-mini) | OpenAI | 28.1% |
| 68 | [GLM-5-Turbo](/models/glm-5-turbo) | Z.AI | 27.8% |
| 69 | [Qwen3.6 Plus](/models/qwen3-6-plus) | Alibaba | 27.8% |
| 70 | [GLM-4.7](/models/glm-4-7) | Z.AI | 27.4% |
| 71 | [Grok 4](/models/grok-4) | xAI | 26.7% |
| 72 | [GPT-5.1-Codex-Max](/models/gpt-5-1-codex-max) | OpenAI | 25.7% |
| 73 | [GPT-5.1-Codex](/models/gpt-5-1-codex) | OpenAI | 25.7% |
| 74 | [GPT-5 (medium)](/models/gpt-5-medium) | OpenAI | 25.4% |
| 75 | [Qwen3.5-122B-A10B](/models/qwen3-5-122b-a10b) | Alibaba | 25.2% |
| 76 | [Qwen3.5-27B](/models/qwen3-5-27b) | Alibaba | 23.9% |
| 77 | [Ling 3.0 Flash](/models/ling-3-0-flash) | InclusionAI | 23.7% |
| 78 | [Ling 3.0 Flash FP8](/models/ling-3-0-flash-fp8) | InclusionAI | 23.7% |
| 79 | [Gemma 4 31B](/models/gemma-4-31b) | Google | 23.6% |
| 80 | [Qwen3.6-27B](/models/qwen3-6-27b) | Alibaba | 23.1% |
| 81 | [Gemini 2.5 Pro](/models/gemini-2-5-pro) | Google | 22.5% |
| 82 | [Qwen3.6-35B-A3B](/models/qwen3-6-35b-a3b) | Alibaba | 22.2% |
| 83 | [MiMo-V2-Omni](/models/mimo-v2-omni) | Xiaomi | 22.1% |
| 84 | [Muse Glimmer 30B](/models/muse-glimmer-30b) | Meta | 22.0% |
| 85 | [Ling 3.0 Flash VL](/models/ling-3-0-flash-vl) | InclusionAI | 22.0% |
| 86 | [Step 3.7 Flash](/models/step-3-7-flash) | StepFun | 21.4% |
| 87 | [Qwen3.5-35B-A3B](/models/qwen3-5-35b-a3b) | Alibaba | 21.0% |
| 88 | [Nemotron 3 Super 100B](/models/nemotron-3-super-100b) | NVIDIA | 20.8% |
| 89 | [o3](/models/o3) | OpenAI | 20.1% |
| 90 | [Qwen3.5 397B (Reasoning)](/models/qwen3-5-397b-reasoning) | Alibaba | 19.8% |
| 91 | [Qwen3.5 397B](/models/qwen3-5-397b) | Alibaba | 19.8% |
| 92 | [GPT-OSS 120B](/models/gpt-oss-120b) | OpenAI | 19.6% |
| 93 | [Gemma 4 26B A4B](/models/gemma-4-26b-a4b) | Google | 19.3% |
| 94 | [Grok 4.1 Fast (Reasoning)](/models/grok-4-1-fast-reasoning) | xAI | 19.3% |
| 95 | [Claude Opus 4.6](/models/claude-opus-4-6) | Anthropic | 19.1% |
| 96 | [Grok 4 Fast (Reasoning)](/models/grok-4-fast-reasoning) | xAI | 19.1% |
| 97 | [Gemini 3.5 Flash-Lite](/models/gemini-3-5-flash-lite) | Google | 18.8% |
| 98 | [Quasar 438B](/models/quasar-438b) | Multiverse Computing | 18.7% |
| 99 | [K-EXAONE 2.0](/models/k-exaone-2-0) | LG AI Research | 18.6% |
| 100 | [GLM-5V-Turbo](/models/glm-5v-turbo) | Z.AI | 17.1% |
| 101 | [Trinity-Large-Preview](/models/trinity-large-preview) | Arcee AI | 15.8% |
| 102 | [DeepSeek-R1](/models/deepseek-r1) | DeepSeek | 15.8% |
| 103 | [Trinity-Large-Thinking](/models/trinity-large-thinking) | Arcee AI | 15.8% |
| 104 | [Gemma 4 12B](/models/gemma-4-12b) | Google | 15.7% |
| 105 | [Gemini 3 Flash](/models/gemini-3-flash) | Google | 15.0% |
| 106 | [DeepSeek V3.1 (Reasoning)](/models/deepseek-v3-1-reasoning) | DeepSeek | 14.3% |
| 107 | [K-Exaone](/models/k-exaone) | LG AI Research | 13.9% |
| 108 | [Mistral Medium 3.5 128B](/models/mistral-medium-3-5-128b) | Mistral | 13.8% |
| 109 | [Claude Sonnet 4.6](/models/claude-sonnet-4-6) | Anthropic | 13.3% |
| 110 | [Claude Opus 4.5](/models/claude-opus-4-5) | Anthropic | 13.2% |
| 111 | [Claude 4.1 Opus Thinking](/models/claude-4-1-opus-thinking) | Anthropic | 12.5% |
| 112 | [Command A+](/models/command-a-plus) | Cohere | 12.0% |
| 113 | [Qwen3 Max](/models/qwen3-max) | Alibaba | 11.9% |
| 114 | [Nemotron 3 Nano 30B](/models/nemotron-3-nano-30b) | NVIDIA | 11.4% |
| 115 | [DeepSeek V3.2](/models/deepseek-v3-2) | DeepSeek | 11.2% |
| 116 | [Granite 4.2 30B](/models/granite-4-2-30b) | IBM | 11.2% |
| 117 | [North Mini Code](/models/north-mini-code-1-0) | Cohere | 11.1% |
| 118 | [GPT-OSS 20B](/models/gpt-oss-20b) | OpenAI | 11.0% |
| 119 | [Sarvam 105B](/models/sarvam-105b) | Sarvam | 11.0% |
| 120 | [Nemotron 3.5 Lightning 30B A3B NVFP4](/models/nemotron-3-5-lightning-30b-a3b-nvfp4) | NVIDIA | 10.6% |
| 121 | [Solar Pro 3](/models/solar-pro-3) | Upstage | 10.3% |
| 122 | [Mistral Small 4](/models/mistral-small-4) | Mistral | 9.9% |
| 123 | [Mistral Small 4 (Reasoning)](/models/mistral-small-4-reasoning) | Mistral | 9.9% |
| 124 | [Granite 4.2 8B](/models/granite-4-2-8b) | IBM | 9.7% |
| 125 | [Ling 3.0 Tiny](/models/ling-3-0-tiny) | InclusionAI | 9.3% |
| 126 | [MiniCPM5-2B](/models/minicpm5-2b) | OpenBMB | 8.9% |
| 127 | [MiMo-V2-Flash](/models/mimo-v2-flash) | Xiaomi | 8.6% |
| 128 | [Grok Code Fast 1](/models/grok-code-fast-1) | xAI | 8.0% |
| 129 | [o3-mini](/models/o3-mini) | OpenAI | 7.9% |
| 130 | [Sarvam 30B](/models/sarvam-30b) | Sarvam | 7.5% |
| 131 | [Qwen3-Omni-30B-A3B-Thinking](/models/qwen3-omni-30b-a3b-thinking) | Alibaba | 7.5% |
| 132 | [Nemotron Ultra 253B](/models/nemotron-ultra-253b) | NVIDIA | 7.4% |
| 133 | [Kimi K2](/models/kimi-k2) | Moonshot AI | 7.4% |
| 134 | [o1](/models/o1) | OpenAI | 7.0% |
| 135 | [GLM-4.5-Air](/models/glm-4-5-air) | Z.AI | 7.0% |
| 136 | [LFM2.5-8B-A1B](/models/lfm2-5-8b-a1b) | LiquidAI | 6.9% |
| 137 | [Celeris-1](/models/celeris-1) | Celeris | 6.8% |
| 138 | [DeepSeek V3.1](/models/deepseek-v3-1) | DeepSeek | 6.7% |
| 139 | [Granite 4.2 3B](/models/granite-4-2-3b) | IBM | 6.6% |
| 140 | [Granite-4.0-H-350M](/models/granite-4-0-h-350m) | IBM | 6.4% |
| 141 | [Ling 2.6 Flash](/models/ling-2-6-flash) | InclusionAI | 6.3% |
| 142 | [LFM2.5-2.6B](/models/lfm2-5-2-6b) | LiquidAI | 6.2% |
| 143 | [Exaone 4.0 1.2B](/models/exaone-4-0-1-2b) | LG AI Research | 5.7% |
| 144 | [GLM-4.6](/models/glm-4-6) | Z.AI | 5.5% |
| 145 | [Granite-4.0-350M](/models/granite-4-0-350m) | IBM | 5.5% |
| 146 | [Grok 4.1 Fast](/models/grok-4-1-fast) | xAI | 5.1% |
| 147 | [LFM2.5-VL-1.6B-Extract](/models/lfm2-5-vl-1-6b-extract) | LiquidAI | 5.1% |
| 148 | [Exaone 4.0 32B](/models/exaone-4-0-32b) | LG AI Research | 5.0% |
| 149 | [Granite-4.0-H-1B](/models/granite-4-0-h-1b) | IBM | 5.0% |
| 150 | [GPT-4.1 mini](/models/gpt-4-1-mini) | OpenAI | 5.0% |
| 151 | [Phi-4 Multimodal Instruct](/models/phi-4-multimodal-instruct) | Microsoft | 5.0% |
| 152 | [Llama 4 Maverick](/models/llama-4-maverick) | Meta | 4.9% |
| 153 | [Nemotron 3 Nano Omni 30B A3B](/models/nemotron-3-nano-omni-30b-a3b) | NVIDIA | 4.8% |
| 154 | [Gemma 4 E2B](/models/gemma-4-e2b) | Google | 4.8% |
| 155 | [Granite-4.0-1B](/models/granite-4-0-1b) | IBM | 4.8% |
| 156 | [Gemini 2.5 Flash](/models/gemini-2-5-flash) | Google | 4.7% |
| 157 | [Qwen3-Omni-30B-A3B-Instruct](/models/qwen3-omni-30b-a3b-instruct) | Alibaba | 4.6% |
| 158 | [Gemini 1.5 Pro](/models/gemini-1-5-pro) | Google | 4.6% |
| 159 | [DeepSeek R1 Distill Qwen 32B](/models/deepseek-r1-distill-qwen-32b) | DeepSeek | 4.6% |
| 160 | [Gemma 3 27B](/models/gemma-3-27b) | Google | 4.4% |
| 161 | [Claude 4 Sonnet](/models/claude-4-sonnet) | Anthropic | 4.3% |
| 162 | [Mistral Large 3](/models/mistral-large-3) | Mistral | 4.2% |
| 163 | [GPT-4.1](/models/gpt-4-1) | OpenAI | 4.2% |
| 164 | [GPT-4o mini](/models/gpt-4o-mini) | OpenAI | 4.2% |
| 165 | [Gemini 1.0 Pro](/models/gemini-1-0-pro) | Google | 4.2% |
| 166 | [Mistral Medium 3](/models/mistral-medium-3) | Mistral | 4.1% |
| 167 | [Claude 3 Haiku](/models/claude-3-haiku) | Anthropic | 4.1% |
| 168 | [Llama 3.1 405B](/models/llama-3-1-405b) | Meta | 4.0% |
| 169 | [Gemma 4 E4B](/models/gemma-4-e4b) | Google | 3.8% |
| 170 | [Llama 4 Scout](/models/llama-4-scout) | Meta | 3.8% |
| 171 | [Phi-4](/models/phi-4) | Microsoft | 3.8% |
| 172 | [GPT-4.1 nano](/models/gpt-4-1-nano) | OpenAI | 3.8% |
| 173 | [Solar Pro 2](/models/solar-pro-2) | Upstage | 3.7% |
| 174 | [Ultravox v0.6 Llama 3.3 70B](/models/ultravox-v0-6-llama-3-3-70b) | Fixie AI | 3.6% |
| 175 | [Qwen2.5 Coder 32B Instruct](/models/qwen2-5-coder-32b-instruct) | Alibaba | 3.5% |
| 176 | [Mistral Large 2](/models/mistral-large-2) | Mistral | 3.3% |
| 177 | [Nova Pro](/models/nova-pro) | Amazon | 3.2% |
| 178 | [GPT-4 Turbo](/models/gpt-4-turbo) | OpenAI | 3.1% |
| 179 | [DeepSeek V3](/models/deepseek-v3) | DeepSeek | 2.9% |
| 180 | [Claude 3 Opus](/models/claude-3-opus) | Anthropic | 2.8% |
| 181 | [GPT-4o](/models/gpt-4o) | OpenAI | 2.4% |

## FAQ

### What does AA-HLE measure?

A display-only Artificial Analysis Humanity's Last Exam score.

### Which model scores highest on AA-HLE?

Claude Fable 5.1 by Anthropic currently leads with a score of 59.1% on AA-HLE.

### How many models are evaluated on AA-HLE?

181 AI models have been evaluated on AA-HLE on BenchLM.

### Does AA-HLE affect BenchLM's overall score?

Not directly. AA-HLE is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Compare Top Models on AA-HLE

- [Claude Fable 5.1 vs Claude Fable 5](/compare/claude-fable-vs-claude-fable-5-1)
- [Claude Fable 5 vs Claude Opus 5](/compare/claude-fable-vs-claude-opus-5)
- [Claude Opus 5 vs GPT-6 Astra](/compare/claude-opus-5-vs-gpt-6-astra)
- [GPT-6 Astra vs GPT-5.6 Sol](/compare/gpt-5-6-sol-vs-gpt-6-astra)