MultiHaystack: leaderboard
Metric: Answer accuracy (%) on all 747 questions, where the model answers from the top-5 items that the E5-V retriever returns from MultiHaystack's 46,260-item mixed corpus of images, videos and documents, graded correct or incorrect by a GPT-4o-mini judge; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 51.41 | #91 |
| 2 | Gemini 2.5 Flash | 43.64 | #237 |
| 3 | Qwen 2 VL 7B | 17.54 | #741 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=multihaystack · How It Works · Data refreshed daily, snapshot 2026-10-11.