PhotoBench (Caption Text Embedding): leaderboard
Metric: Recall@10 (%) on the Chinese queries, averaged over the three albums, where GPT-4o first captions every photo and the model ranks photos by text-to-text similarity between the query and the captions; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen3-Embedding-8B | 51.7 | #674 |
| 2 | Qwen3-Embedding-0.6B | 48.4 | #788 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=photobench-caption-text-embedding · How It Works · Data refreshed daily, snapshot 2026-10-11.