Baidu Open-Sources Unlimited OCR 2 for Hundred-Page Document Parsing
Baidu has released Unlimited OCR 2, an open-source OCR model built for very long documents. The 3B model parses books, filings, and reports of hundreds of pages in a single request. Its core change is SR-SWA, a sparse attention scheme that keeps per-step decoding cost nearly flat. Weights ship under the MIT license, with code on GitHub.
TL;DR
- Size: 3.34B parameters. Active parameters not disclosed. Maximum position length of 131,072 tokens.
- Runs on: A single CUDA GPU in BF16. Baidu tested on an NVIDIA A800-80GB. BF16 weights total about 6.7 GB.
- Performance: First place on both multi-page benchmarks Baidu tested, against nine other models (self-reported).
- Best result: 77.1 overall on MarathonDocBench (hundreds of pages). The next model, Kimi-K3, scores 33.7.
- Weakest result: 48.2 on heading hierarchy (TEDS) on MarathonDocBench, its lowest sub-score.
What Is Unlimited OCR 2?
Unlimited OCR 2 is a 3B vision-language model from Baidu that turns ordered page images into structured text. Output covers text, tables, formulas, figures, headings, and reading order. It succeeds Unlimited-OCR, released in June 2026 to push DeepSeek-OCR further.
The release lands as long-context document AI becomes a practical requirement rather than a research curiosity: enterprises in 2026 increasingly feed entire contracts, regulatory filings, and scanned archives into single-model pipelines, where context-window efficiency—not raw accuracy on single pages—determines feasibility.
via MarkTechPost