CorpusQA (128K): leaderboard
Metric: Accuracy (%; mean of the four domain accuracies, 128K-token corpora, DeepSeek-V3 judge). Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 82.02 |
| 2 | Gemini 2.5 Pro | 80.19 |
| 3 | GPT-5 Mini | 79.88 |
| 4 | Qwen 3 235B A22B 2507 (Thinking) | 78.08 |
| 5 | Gemini 2.5 Flash | 77.81 |
| 6 | DeepSeek R1 0528 | 74.72 |
| 7 | MiniMax M1 80k | 71.02 |
| 8 | Qwen 3 30B A3B 2507 (Thinking) | 69.59 |
Interactive version: theaggregate.ai/benchmark?slug=corpusqa-128k · How It Works · Data refreshed daily, snapshot 2026-09-25.