SIN-Bench: leaderboard
Metric: Average Overall Score (%; mean of the SIN-Find, SIN-Verify, SIN-QA and SIN-Summary task scores). Source: arxiv.org. Saturation forecast: Around July 2028. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 56.6 |
| 2 | Claude Sonnet 4.5 (Thinking) | 54.9 |
| 3 | GPT-5 | 54.4 |
| 4 | Gemini 2.5 Pro | 51 |
| 5 | Grok 4 | 49.5 |
| 6 | Qwen 3 VL 8B (Thinking) | 45.2 |
| 7 | Qwen 3 VL 30B A3B (Thinking) | 44.8 |
Interactive version: theaggregate.ai/benchmark?slug=sin-bench · How It Works · Data refreshed daily, snapshot 2026-09-25.