BrainBench (EEG) - BrainAgent - Sleep Assessment: leaderboard
BrainBench (arXiv 2608.04156v2, Table 1) score on the Sleep Assessment subset built on clinical sleep recordings, with each model running through the structured BrainAgent tool workflow in an isolated container and its report and artifacts checked by task validators. 0-100; higher is better. Optional reasoning modes disabled.
Metric: EEG analysis score (0-100). Source: arxiv.org. Saturation forecast: Around 2029. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol (Non-reasoning) | 70.56 | #22 (GPT-5.6 Sol) |
| 2 | Qwen 3.7 Max (Non-reasoning) | 70.54 | #85 (Qwen 3.7 Max) |
| 3 | Kimi K2.5 (Non-reasoning) | 68.09 | #152 (Kimi K2.5) |
| 4 | Qwen 3.7 Plus (Non-reasoning) | 68.08 | #98 (Qwen 3.7 Plus) |
| 5 | DeepSeek V4 Flash (Non-reasoning) | 67.97 | #155 (DeepSeek V4 Flash) |
| 6 | Claude Opus 5 | 66.41 | #16 |
| 7 | MiniMax-M2.5 | 66.31 | #316 |
| 8 | GLM-5 (Non-reasoning) | 66.17 | #149 (GLM-5) |
| 9 | GPT-5.6 Terra (Non-reasoning) | 66.06 | #40 (GPT-5.6 Terra) |
| 10 | Qwen 3.5 122B A10B (Non-reasoning) | 65.98 | #184 (Qwen 3.5 122B A10B) |
| 11 | GPT-5.6 Luna (Non-reasoning) | 62.8 | #79 (GPT-5.6 Luna) |
| 12 | Qwen 3.5 35B A3B (Non-reasoning) | 59.35 | #268 (Qwen 3.5 35B A3B) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=brainbench-eeg-brainagent-sleep-assessment · How It Works · Data refreshed daily, snapshot 2026-11-05.