← Benchmarks
Humanity's Last Exam
Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions
id humanity's-last-exam · max 1 · 104 models reported
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.1 Anthropic | 0.65 |
| 2 | Claude Mythos Preview Anthropic | 0.65 |
| 3 | Claude Opus 5 Anthropic | 0.65 |
| 4 | Claude Fable 5 Anthropic | 0.65 |
| 5 | GLM-5.3 Zhipu AI · open | 0.63 |
| 6 | Muse Spark 1.1 Meta | 0.62 |
| 7 | DeepSeek-V4-Pro-0813 DeepSeek · open | 0.60 |
| 8 | Muse Spark Meta | 0.58 |
| 9 | Claude Opus 4.8 Anthropic | 0.58 |
| 10 | Claude Sonnet 5 Anthropic | 0.57 |
| 11 | GPT-5.5 Pro OpenAI | 0.57 |
| 12 | Kimi K3 Moonshot AI · open | 0.56 |
| 13 | Seed 2.1 Pro ByteDance | 0.56 |
| 14 | GLM-5.3-Flash Zhipu AI · open | 0.55 |
| 15 | Claude Opus 4.7 Anthropic | 0.55 |
| 16 | GLM-5.2 Zhipu AI · open | 0.55 |
| 17 | Seed 2.1 Turbo ByteDance | 0.55 |
| 18 | Claude Opus 4.6 Anthropic | 0.53 |
| 19 | GLM-5.1 Zhipu AI · open | 0.52 |
| 20 | GPT-5.5 OpenAI | 0.52 |
| 21 | Gemini 3.1 Pro Google | 0.51 |
| 22 | Kimi K2-Thinking-0905 Moonshot AI · open | 0.51 |
| 23 | Grok-4 Heavy xAI | 0.51 |
| 24 | Kimi K2.5 Moonshot AI · open | 0.50 |
| 25 | Claude Sonnet 4.6 Anthropic | 0.49 |
| 26 | Qwen3.5-27B Alibaba Cloud / Qwen Team · open | 0.48 |
| 27 | DeepSeek-V4-Pro-Max DeepSeek · open | 0.48 |
| 28 | Qwen3.5-122B-A10B Alibaba Cloud / Qwen Team · open | 0.47 |
| 29 | Qwen3.5-35B-A3B Alibaba Cloud / Qwen Team · open | 0.47 |
| 30 | Gemini 3 Pro Google | 0.46 |
| 31 | DeepSeek-V4-Flash-Max DeepSeek · open | 0.45 |
| 32 | Qwen3.8 Max Alibaba Cloud / Qwen Team · open | 0.44 |
| 33 | Gemini 3 Flash Google | 0.43 |
| 34 | GLM-4.7 Zhipu AI · open | 0.43 |
| 35 | Qwen3.7 Max Alibaba Cloud / Qwen Team | 0.41 |
| 36 | DeepSeek-V3.2 DeepSeek · open | 0.41 |
| 37 | DeepSeek-V4-Flash-0423 DeepSeek · open | 0.40 |
| 38 | Gemini 3.5 Flash Google | 0.40 |
| 39 | Grok-4 xAI | 0.40 |
| 40 | GPT-5.4 OpenAI | 0.40 |
| 41 | ERNIE 5.0 Baidu | 0.39 |
| 42 | Nemotron 3 Ultra (550B A55B) NVIDIA · open | 0.37 |
| 43 | DeepSeek-V4.1-Flash DeepSeek · open | 0.37 |
| 44 | GPT-5.2 Pro OpenAI | 0.37 |
| 45 | Kimi K2.6 Moonshot AI · open | 0.36 |
| 46 | Qwen3.8 Flash Alibaba Cloud / Qwen Team | 0.36 |
| 47 | Qwen3.8-Flash-Next Alibaba Cloud / Qwen Team · open | 0.36 |
| 48 | Qwen3.7-Plus Alibaba Cloud / Qwen Team | 0.35 |
| 49 | GPT-5.2 OpenAI | 0.34 |
| 50 | MiMo-V2.5-Pro Xiaomi · open | 0.34 |
| 51 | Inkling-Small Thinking Machines Lab · open | 0.32 |
| 52 | Qwen3.8-27B Alibaba Cloud / Qwen Team · open | 0.31 |
| 53 | DeepSeek-V3.2-Speciale DeepSeek · open | 0.31 |
| 54 | Qwen3.6 Plus Alibaba Cloud / Qwen Team | 0.29 |
| 55 | Qwen3.5-397B-A17B Alibaba Cloud / Qwen Team · open | 0.29 |
| 56 | GPT-5.4 mini OpenAI | 0.28 |
| 57 | Gemma 4 31B Google · open | 0.27 |
| 58 | LongCat-Flash-Thinking-2601 Meituan · open | 0.25 |
| 59 | DeepSeek-V3.2 (Thinking) DeepSeek · open | 0.25 |
| 60 | GPT-5 OpenAI | 0.25 |
| 61 | GPT-5.4 nano OpenAI | 0.24 |
| 62 | Qwen3.6-27B Alibaba Cloud / Qwen Team · open | 0.24 |
| 63 | Nemotron 3 Super (120B A12B) NVIDIA · open | 0.23 |
| 64 | MiMo-V2-Flash Xiaomi · open | 0.22 |
| 65 | MiniMax M2.1 MiniMax · open | 0.22 |
| 66 | Muse Glimmer-30B Meta · open | 0.22 |
| 67 | Gemini 2.5 Pro Preview 06-05 Google | 0.22 |
| 68 | Qwen3.6-35B-A3B Alibaba Cloud / Qwen Team · open | 0.21 |
| 69 | Grok 4 Fast xAI | 0.20 |
| 70 | DeepSeek-V3.2-Exp DeepSeek · open | 0.20 |
| 71 | Qwen3-235B-A22B-Thinking-2507 Alibaba Cloud / Qwen Team · open | 0.18 |
| 72 | MAI-Code-1-Flash Microsoft | 0.18 |
| 73 | Gemini 2.5 Pro Google | 0.18 |
| 74 | DeepSeek-R1-0528 DeepSeek · open | 0.18 |
| 75 | Gemma 4 26B-A4B Google · open | 0.17 |
| 76 | GLM-4.6 Zhipu AI · open | 0.17 |
| 77 | GPT-5 mini OpenAI | 0.17 |
| 78 | Gemini 3.1 Flash-Lite Google | 0.16 |
| 79 | DeepSeek-V3.1 DeepSeek · open | 0.16 |
| 80 | Nemotron 3 Nano (30B A3B) NVIDIA · open | 0.15 |
| 81 | GPT OSS 120B OpenAI · open | 0.15 |
| 82 | o3 OpenAI | 0.15 |
| 83 | o4-mini OpenAI | 0.15 |
| 84 | GLM-4.5 Zhipu AI · open | 0.14 |
| 85 | GLM-4.7-Flash Zhipu AI · open | 0.14 |
| 86 | Qwen3 VL 235B A22B Thinking Alibaba Cloud / Qwen Team · open | 0.14 |
| 87 | MiniMax M2 MiniMax · open | 0.13 |
| 88 | DiffusionGemma 26B-A4B Google · open | 0.12 |
| 89 | Nemotron 3.5 Lightning (30B A3B) NVIDIA · open | 0.12 |
| 90 | Sarvam-105B Sarvam AI · open | 0.11 |
| 91 | Gemini 2.5 Flash Google | 0.11 |
| 92 | GPT OSS 20B OpenAI · open | 0.11 |
| 93 | GLM-4.5-Air Zhipu AI · open | 0.11 |
| 94 | Magistral Medium Mistral AI · open | 0.09 |
| 95 | GPT-5 nano OpenAI | 0.09 |
| 96 | MiniMax M1 80K MiniMax · open | 0.08 |
| 97 | MiniMax M1 40K MiniMax · open | 0.07 |
| 98 | GPT-4.1 OpenAI | 0.05 |
| 99 | GPT-4o OpenAI | 0.05 |
| 100 | Gemma 4 12B Google · open | 0.05 |