Xiaomi open-sources Xiaomi-OCR-0, a 0.8B document understanding model topping sub-1B parsing
gnukeith · x · 2026-10-30
Xiaomi Research released Xiaomi-OCR-0, an open-source document understanding model with only 0.8B parameters that unifies document parsing, key information extraction, and OCR-oriented visual question answering.
- Scores 95.24 on Real5-OmniDocBench and averages 83.2 across five OCR-oriented VQA benchmarks
- Ranks first in document parsing among sub-1B models on OmniDocBench v1.6
- Ships with model weights, inference code, an interactive demo, an MCP service, and an Agent Skill on Hugging Face and GitHub
More from Models
- The meme is real: DeepSeek-led open-weight models hit 21 ops in Q3, beating all prior quarters combined — Jsevillamol · 2026-10-30
- Polymarket puts Google at 59% to lead the AI race by year-end — Polymarket · 2026-10-30
- Nanbeige4.3-33B with 512K context tops agent benchmarks, real-world feedback sought — xdcfret1 · 2026-10-30
- LLMs struggle badly on tasks with more than one right answer — andpoul · 2026-10-30
- Nace AI open-sources Drex-Max v1: 27B multimodal decision model scoring 255 options in one pass — techNmak · 2026-10-30
- 0.87MB Hindi typing model Lekhani beats a 14x larger model in-browser — SuB8u · 2026-10-30