--- format: "aidr-story-markdown/v1" id: "1aee347ea02e0e1c4fa408c194c21fec04e32e3a7816603a116ddacbe26ee38e" canonical_url: "https://aidr.today/1aee347e?lang=en" title: "Baidu Open Sources New OCR Tool That Parses Entire Books in Single Requests" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-27T08:37:45.000Z" category: "Open Source" topics: ["open-source"] source_urls: ["https://huggingnews.com/ai/baidu-open-sources-new-ocr-tool-that-parses-entire-books-in-single-reque-a0dd20fd","https://x.com/Baidu_Inc/status/2114910381311512987","https://x.com/sgl_project/status/2114945367712366718","https://marketbrief.now/ai/baidu-open-sources-new-ocr-tool-that-parses-entire-books-in-single-reque-a0dd20fd"] summary: "Baidu has released Unlimited OCR 2, an open source model that can parse documents of virtually any length while maintaining a stable memory footprint. The model uses a SR-SWA mechanism to condense attention interactions and can process a whole book in a single request with flat memory usage across hundreds of pages. It recorded state-of-the-art performance on the MarathonDocBench and MPDocBench benchmarks." --- # Baidu Open Sources New OCR Tool That Parses Entire Books in Single Requests > [Open the canonical story]() **Published:** 2026-10-27T08:37:45.000Z **Category:** Open Source **Topics:** open\-source ## Summary Baidu has released Unlimited OCR 2, an open source model that can parse documents of virtually any length while maintaining a stable memory footprint\. The model uses a SR\-SWA mechanism to condense attention interactions and can process a whole book in a single request with flat memory usage across hundreds of pages\. It recorded state\-of\-the\-art performance on the MarathonDocBench and MPDocBench benchmarks\. ## Sources - [Story source]() - [Story source]() - [Supporting source]() - [Story source]()