SuperCLUE-LongContext - 256K Overall: leaderboard
SuperCLUE is the Chinese-language evaluation programme run by CLUE; its specialised boards each cover one professional domain and are graded by the organisers rather than self-reported. This board covers long context at a 256K window.
Metric: Score. Source: www.superclueai.com. Saturation forecast: Around October 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 (High) | 98.24 | #31 (GPT-5.5) |
| 2 | Gemini 3.1 Pro (Preview) (High) | 85.43 | #58 (Gemini 3.1 Pro (Preview)) |
| 3 | Qwen 3.7 Max (Thinking) | 76.2 | #75 (Qwen 3.7 Max) |
| 4 | DeepSeek V4 Pro (Max) | 74.54 | #102 (DeepSeek V4 Pro) |
| 5 | Kimi K3 (Max) | 68.09 | #30 (Kimi K3) |
| 6 | Kimi K2.6 | 56.14 | #104 |
| 7 | Doubao-Seed-2-0-Pro-260215 (High) | 55.1 | |
| 8 | GLM-5.2 (Max) | 53.71 | #73 (GLM-5.2) |
| 9 | GLM-5.1 | 51.53 | #97 |
| 10 | MiMo-V2.5-Pro | 45.34 | #158 |
| 11 | MiniMax-M3 | 28.04 | #136 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=superclue-longcontext-256k-overall · How It Works · Data refreshed daily, snapshot 2026-10-20.