PortBench-QA: leaderboard
Metric: Mean item score (%; mean of seven QA templates, 50 test questions per model per template, thinking on, temperature 0). Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.7 Max (Thinking) | 81.9 |
| 2 | DeepSeek V4 Flash (Thinking) | 81.9 |
| 3 | DeepSeek V4 Pro (Thinking) | 81.8 |
| 4 | Qwen 3.6 Plus (Thinking) | 78.3 |
| 5 | GLM-5.1 (Thinking) | 75.7 |
| 6 | Qwen 3.6 35B A3B (Thinking) | 68.4 |
| 7 | Kimi K2.6 (Thinking) | 51.1 |
Interactive version: theaggregate.ai/benchmark?slug=portbench-qa · How It Works · Data refreshed daily, snapshot 2026-09-26.