OneSearch-VL: a unified multimodal deep research agent for images and video
A new arXiv paper introduces OneSearch-VL, which centers on a Visually Grounded Evidence Graph (VGEG) that links localized visual anchors, entity relations, source-supported facts and answer-producing operations so that single-image, multi-image and video deep research can share one workflow of grounding, retrieval and fact composition. The authors use VGEG to build the OneSearch-VL-SFT-110K and OneSearch-VL-RL-10K training sets, an evidence-aware reward (EVGR) for RL, plus the OneSearch-MI-Bench and OneSearch-Video-Bench evaluations. They report that OneSearch-VL-8B beats Qwen3-VL-8B with tool access by 20.2 and 17.6 percentage points on the two new benchmarks, and also gains across 7 image benchmarks and VideoDR.
It offers a shared evidence-graph framework for training and evaluating multimodal deep-research agents across image and video tasks.
Computer Science > Computer Vision and Pattern Recognition
arXiv:2610.12419 (cs)
[Submitted on 8 Oct 2026]
Title:OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video
Authors:Hongyu Li, Manyuan Zhang, Kaituo Feng, Shu Chen, Dian Zheng, Hao Li, Hao Yu, Zhangquan Chen, Zoey Guo, Ray Zhang, Shaofei Huang, Tianrui Hui, Linjiang Huang, Si Liu
View a PDF of the paper titled OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video, by Hongyu Li and 13 other authors
#OneSearch-VL#OneSearch-VL-8B#Qwen3-VL#OneSearch-MI-Bench#OneSearch-Video-Bench