Baidu Open-Sources Long-Doc OCR Model
brianrkelly · x · 2026-07-20
Baidu has open-sourced an OCR model named Unlimited-OCR. With only 3 billion parameters, it is designed to read entire long documents in one go, claiming the ability to directly parse 100-page PDFs. It supports 32K context, multiple languages, and can run on local hardware.\n\nThe post also shares several results: it achieves 93% on standard parsing benchmarks, 6 points higher than the baseline, and for documents over 40 pages, the error rate drops below 0.11. The model is compatible with Transformers, vLLM, SGLang, Docker, Ollama, and llama.cpp. The author notes it has reached 1.9 million downloads on Hugging Face and was developed to push the boundaries of DeepSeek-OCR even further.
Related event: Baidu Open-Sources Unlimited-OCR for Long Documents(4 posts)→
More from Models
- GPT 6.1 Terra vs Claude Sonnet 5.5: speed and value vs raw capability — teortaxesTex · 2026-10-27
- Harmonic Teases: AI Taught to Prove Math, 'Something Harder' Coming Soon — satnam6502 · 2026-10-27
- OpenAI quietly swaps Plus/Pro default to cheaper GPT-6.1 Terra and cuts Codex allowance — kimmonismus · 2026-10-27
- 'Never been this uninterested after an OpenAI model release' — theteknosaur · 2026-10-27
- Dev questions reports of Chinese models writing insecure code, citing thin methods and conflicts — dbreunig · 2026-10-27
- Aidan Gomez: LLM probing outside servers was out-of-bounds, not proof of intent — AccBalanced · 2026-10-27