Two Pure-Vision Pipelines at the Top of ViDoRe V3
Blog post from Hugging Face
Two open-weight, image-only retrieval pipelines achieved leading results on the ViDoRe V3 benchmark, which evaluates complete document retrieval systems across eight English visual-document datasets. Both use Tencent’s EVIE-4.5B late-interaction page-image retriever as a first stage: VisAgentRet combines it with GLM-5.3-Flash, which iteratively reformulates queries, searches, and inspects candidate pages, reaching an average nDCG@10 of 70.01 but requiring about 49.3 seconds per query; the EVIE-4.5B plus JEV-27B-VL pipeline reranks the top 16 page images in a single multiple-choice inference pass, attaining 68.25 nDCG@10 at 1.47 seconds per query. EVIE-4.5B alone scored 66.76, while the agent and reranker improved results on most datasets, though physics remained stronger with the base retriever. A comparison of agent inputs found that viewing page images slightly outperformed OCR-extracted text, 70.01 versus 69.69, while using fewer prompt tokens. The authors released code, model configurations, evaluation commands, and hardware details under Apache-2.0 licenses to support reproduction.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| Jev | 2 | 1,284 | 96 | 54 | -40% |
| Vector Search | 2 | 1,978 | 415 | 135 | +18% |
| LLM | 1 | 6,578 | 1,243 | 236 | +9% |
| MCP | 1 | 11,921 | 1,232 | 235 | +12% |
| Web search for AI agents | 1 | 22 | 13 | 9 | +100% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.