DR Pixel, an open-source deep research agent, reasons over page images without OCR

AITiansheng Hu introduces DR Pixel, an open-source deep research agent that searches and reasons directly over page images without OCR. On 1.06M pages, the authors' trained 27B model matches GPT-5.4, according to the post. The work is joint research from BerkeleyNLP and rtnldynamics, with the paper linked at arXiv 2611.03097.

Tiansheng Hu

@Hughie_Hu43

View on X

(1/7) 🚀 Introducing DR Pixel, the first open-source deep research agent that searches and reasons directly over page images, no OCR. On 1.06M pages, our trained 27B model matches GPT-5.4.

📄 https://arxiv.org/abs/2611.03097
Joint work from @BerkeleyNLP, @rtnldynamics!

Source details

Tiansheng Hu · x.comPublished Nov 3, 8:27 AM ET · added here 1:19 PM ET

Original title: (1/7) 🚀 Introducing DR Pixel, the first open-source deep research agent that searches and reasons directly over page images, no OCR. On ...

Topics

AI agents Multimodal AI Open-source ecosystem Papers & research Evals & benchmarks