{"full_name":"docling-project/docling","owner":"docling-project","name":"docling","url":"https://github.com/docling-project/docling","description":"Get your documents ready for gen AI","homepage":"https://docling-project.github.io/docling","topics":[],"language":"Python","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/188446108?v=4","created_at":"2024-07-09T07:50:26Z","pushed_at":"2026-11-10T07:57:08Z","metrics":{"stars":70356,"forks":5306,"watchers":224,"issues":2346,"prs":2351,"releases":238,"issues_open":null,"issues_closed":null,"prs_open":null,"prs_closed":null,"prs_merged":null,"discussions":483},"growth":{"day":{"stars":41,"forks":8,"base_day":"2026-11-09"},"week":{"stars":341,"forks":54,"base_day":"2026-11-03"},"month":{"stars":1943,"forks":303,"base_day":"2026-10-11"}},"history":{"dates":["2026-09-29","2026-09-30","2026-10-01","2026-10-02","2026-10-03","2026-10-04","2026-10-05","2026-10-06","2026-10-07","2026-10-08","2026-10-09","2026-10-10","2026-10-11","2026-10-12","2026-10-13","2026-10-14","2026-10-15","2026-10-16","2026-10-17","2026-10-18","2026-10-19","2026-10-20","2026-10-21","2026-10-22","2026-10-23","2026-10-24","2026-10-25","2026-10-26","2026-10-27","2026-10-28","2026-10-29","2026-10-30","2026-10-31","2026-11-01","2026-11-02","2026-11-03","2026-11-04","2026-11-09","2026-11-10"],"stars":[66987,67522,67621,67702,67856,67952,67991,68128,68180,68265,68309,68334,68413,68455,68494,68543,68597,68644,68761,68820,68884,68933,68987,69012,69122,69203,69279,69374,69453,69508,69627,69684,69752,69817,69878,70015,70073,70315,70356],"forks":[4833,4875,4888,4899,4917,4925,4932,4948,4956,4978,4984,4991,5003,5007,5017,5025,5030,5035,5056,5068,5078,5089,5098,5101,5110,5121,5133,5152,5164,5170,5190,5202,5204,5214,5220,5252,5264,5298,5306],"watchers":[227,227,227,227,228,228,228,229,230,230,229,229,228,228,228,228,228,229,229,229,229,230,230,230,229,229,228,228,229,224,225,225,224,225,225,224,224,224,224],"issues_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null],"prs_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null]},"boards":{},"same_owner":[],"issues":[{"number":3910,"title":"Docling and llm granite-docling 258m (RockMan256) with LM Studio-0.4.20","comments":53,"created_at":"2026-07-29T12:18:28Z"},{"number":287,"title":"Identify table of contents for better chunking Hierarchy Identification","comments":44,"created_at":"2024-11-09T19:55:13Z"},{"number":1023,"title":"Export to markdown only contains H2 headers","comments":26,"created_at":"2025-02-19T16:45:09Z"},{"number":2102,"title":"Please add more descriptions on how to use VLM.","comments":23,"created_at":"2025-08-19T12:06:28Z"},{"number":2779,"title":"Docling consumes all available memory and gets killed","comments":19,"created_at":"2025-12-13T12:21:00Z"},{"number":1111,"title":"Convert PDF file error [RuntimeError: Invalid code point]","comments":18,"created_at":"2025-03-04T09:50:47Z"},{"number":3201,"title":"SIGABRT crash on Python 3.12 — all docling versions fail on Databricks Serverless","comments":17,"created_at":"2026-03-27T19:16:14Z"},{"number":2186,"title":"Using vision model for document conversion, clarity about picture and table enrichment","comments":16,"created_at":"2025-09-04T01:23:59Z"}],"discussions":[{"number":5239,"title":"Would Docling be a named pilot for an open tool that records why AI-generated code changed?","comments":0,"category":"General","created_at":"2026-11-08T04:34:29Z"},{"number":5093,"title":"Proposal: alignment of independently converted document revisions","comments":0,"category":"Ideas","created_at":"2026-10-29T15:15:15Z"},{"number":5074,"title":"docmutant: real-document evidence for parser bugs, without sharing the document","comments":0,"category":"Show and tell","created_at":"2026-10-28T13:58:52Z"},{"number":5066,"title":"Page numbers and bounding boxes for DOCX: filling `prov` in Docling's own output from a Word-compatible layout engine","comments":0,"category":"Show and tell","created_at":"2026-10-28T04:12:35Z"},{"number":4997,"title":"Any suggestions for the best approach to read and extract from pdfs and in a great accuracy and low cost? A solution was about deploying gemma model on a rented gpu on amazon with jevv as router.. any recommendations ?","comments":2,"category":"Q&A","created_at":"2026-10-23T22:04:20Z"},{"number":4883,"title":"Docling CLI, Serve and MCP tools included in AgentManager","comments":0,"category":"Show and tell","created_at":"2026-10-18T11:50:15Z"},{"number":4837,"title":"List Item Serialization Policy: Data Model & Serialization Rules","comments":0,"category":"General","created_at":"2026-10-14T21:11:35Z"},{"number":4835,"title":"Integrating Docling into a Production RAG Pipeline: Table Handling, Dual Representations, and VLM Fallbacks","comments":1,"category":"Show and tell","created_at":"2026-10-14T18:43:46Z"}],"latest_release":{"name":null,"tag":"v2.149.0","published_at":"2026-11-08T08:25:48Z","body":"### Feature\n\n* **pdf:** Add opt-in embedded PDF image extraction ([#5143](https://github.com/docling-project/docling/issues/5143)) ([`3df6c93`](https://github.com/docling-project/docling/commit/3df6c9318fc6d91fde424c90a28fc7d7e5f080dc))\n* Refactored the cli to be more intuitive ([#5214](https://github.com/docling-project/docling/issues/5214)) ([`55f21d6`](https://github.com/docling-project/docling/commit/55f21d6a439a236965542f4fd2f21d8366306b93))\n* **asr:** Add support for mlx-audio ([#5181](https://github.com/docling-project/docling/issues/5181)) ([`756059c`](https://github.com/docling-project/docling/commit/756059c2666753890e3f531ff5d6416be8b3df0e))\n* **vlm:** Add MiniCPM-OCR preset ([#5134](https://github.com/docling-project/docling/issues/5134)) ([`1282e3d`](https://github.com/docling-project/docling/commit/1282e3d9fa34f28772785177f1f330936e8df96a))\n\n### Fix\n\n* **md:** Decode only terminated character references in text ([#5231](https://github.com/docling-project/docling/issues/5231)) ([`414d669`](https://github.com/docling-project/docling/commit/414d669f81646eb4548a66dd2ec3869c56a075ca))\n* **latex:** Number the items of enumerate lists ([#4518](https://github.com/docling-project/docling/issues/4518)) ([`1d9a79c`](https://github.com/docling-project/docling/commit/1d9a79c85028d4a2917e8026de338158f5d9fa88))\n* **md:** Keep backslash escapes inside the surrounding text ([#4991](https://github.com/docling-project/docling/issues/4991)) ([`bcb3678`](https://github.com/docling-project/docling/commit/bcb3678fad7426549643f54c5eceab77925ac5c1))\n* **jats:** Keep tables published as an image ([#5078](https://github.com/docling-project/docling/issues/5078)) ([`1e26b84`](https://github.com/docling-project/docling/commit/1e26b84c5f5690128ddcd65a20a5d7ffb9e80dcd))\n* **asciidoc:** Fill a table row from the column count of the first line ([#5010](https://github.com/docling-project/docling/issues/5010)) ([`26cff4d`](https://github.com/docling-project/docling/commit/26cff4d67939ebd33f19594acdb53ad2c5a531dc))\n* **reading-order:** Skip a horizontal dilation that overlaps another element ([#5131](https://github.com/docling-project/docling/issues/5131)) ([`55df934`](https://github.com/docling-project/docling/commit/55df93475b4a5e6613aab95ede217308acf870c0))\n* **html:** Keep an image caption with its picture ([#5037](https://github.com/docling-project/docling/issues/5037)) ([`0ae7561`](https://github.com/docling-project/docling/commit/0ae75619c886106f49a9e14202ce37741e938a0f))\n* Keep pictures whose mode cannot be written as PNG ([#4683](https://github.com/docling-project/docling/issues/4683)) ([`1331d70`](https://github.com/docling-project/docling/commit/1331d70c7c09c83bc594dbb00e831fb564134e15))\n* **html:** Keep sibling code tags of pre in one code block ([#4998](https://github.com/docling-project/docling/issues/4998)) ([`ec9911e`](https://github.com/docling-project/docling/commit/ec9911e29c9c1f5309280698b0dba17d5e002c36))\n* **jats:** Preserve table footnotes ([#4818](https://github.com/docling-project/docling/issues/4818)) ([`c715294`](https://github.com/docling-project/docling/commit/c715294cd774ef0a59dd8e5c8a63ae3432cd611e))\n* **csv:** Read Excel's sep= first line as the delimiter ([#5200](https://github.com/docling-project/docling/issues/5200)) ([`1436f13`](https://github.com/docling-project/docling/commit/1436f130d254833863b2fab9ad2d97946d7a6d1a))\n* **uspto:** Map ISO 8879 &sgr; to sigma and keep &sfgr; ([#5202](https://github.com/docling-project/docling/issues/5202)) ([`4e57113`](https://github.com/docling-project/docling/commit/4e571139022e886cacc01a113ff695f2d57d790d))\n\n### Documentation\n\n* Fix VLM configuration in GPU guide ([#5102](https://github.com/docling-project/docling/issues/5102)) ([`98dfa0f`](https://github.com/docling-project/docling/commit/98dfa0f99c223d09e262e7629e2f4076482e27f7))\n* Fix spelling of 'preferred' in docstrings ([#5213](https://github.com/docling-project/docling/issues/5213)) ([`d7654f4`](https://github.com/doclin"},"readme":true,"readme_zh":false,"news":[{"title":"Docling-serve: Running Docling as an API service","url":"https://github.com/docling-project/docling-serve","hn":"https://news.ycombinator.com/item?id=45090760","source":"Hacker News","lang":"en","date":"2025-09-01","points":3,"comments":0},{"title":"Docling simplifies document processing, parsing diverse formats","url":"https://github.com/docling-project/docling","hn":"https://news.ycombinator.com/item?id=45173954","source":"Hacker News","lang":"en","date":"2025-09-08","points":2,"comments":0},{"title":"Docling: Document Processing, Diverse Format Parsing, Advanced PDF Understanding","url":"https://github.com/docling-project/docling","hn":"https://news.ycombinator.com/item?id=43993402","source":"Hacker News","lang":"en","date":"2025-05-15","points":2,"comments":0},{"title":"Docling - Composite file format","url":"https://github.com/docling-project/docling","hn":"https://news.ycombinator.com/item?id=48946749","source":"Hacker News","lang":"en","date":"2026-07-17","points":1,"comments":0},{"title":"Docling - Get your documents ready for gen AI","url":"https://github.com/docling-project/docling","hn":"https://news.ycombinator.com/item?id=46262544","source":"Hacker News","lang":"en","date":"2025-12-14","points":1,"comments":0},{"title":"Office 文档怎么喂给 AI？Docling 实测","url":"https://www.bilibili.com/video/BV1XnoVBvEuY/","source":"bilibili","lang":"zh","date":"2026-04-23"},{"title":"从文档到AI：docling一站式数据预处理流水线详解","url":"https://blog.csdn.net/gitblog_00897/article/details/151147797","source":"blog.csdn","lang":"zh","date":"2026-01-05"},{"title":"【Githubの宝石シリーズ】Docling: 生成AI向けドキュメント処理を革新するオープンソースライブラリ","url":"https://note.com/bright_hosta5/n/n5dadec22ee2b","source":"note","lang":"zh","date":"2025-10-05"},{"title":"IBM macht Dokumente schlau: Granite-Docling verwandelt PDFs in KI-Futter","url":"https://www.it-daily.net/it-management/ki/ibm-granite-docling","source":"it-daily.net","lang":"en","date":"2025-09-28"},{"title":"IBM Releases New Granite-Docling Model to Deliver End-to-End Document Understanding","url":"https://www.dbta.com/Editorial/News-Flashes/IBM-Releases-New-Granite-Docling-Model-to-Deliver-End-to-End-Document-Understanding-171525.aspx","source":"dbta.com","lang":"en","date":"2025-09-21"},{"title":"【GitHub项目推荐--Docling：面向生成式AI的智能文档处理框架】","url":"https://blog.csdn.net/j8267643/article/details/151676285","source":"blog.csdn","lang":"zh","date":"2025-09-13"},{"title":"Docling: An open-source tool kit for advanced document processing","url":"https://www.infoworld.com/article/3997240/docling-an-open-source-tool-kit-for-advanced-document-processing.html","source":"InfoWorld","lang":"en","date":"2025-05-29"},{"title":"Datalab Lift vs the Field: How a 9B Schema-First Extractor Compares with NuExtract3, LlamaExtract, Marker, and Docling","url":"https://news.google.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?oc=5","source":"MarkTechPost","lang":"en","date":"2026-07-09"},{"title":"IBM Granite-Docling：通过一个微型模型实现端到端文档理解","url":"https://news.google.com/rss/articles/CBMilAFBVV95cUxOei1yNUR4NVU3d3M5V1NneU0xbklpdXREZUdHek1NOW81djloWk9zQlkydE1Yclk0UVIzZ056c3lJb2pDNFlkaHV2NUI3MEJPQ01IM3Y3cG4xbk9NRVpaOUZsVjVpWmNxN3BrWEpGbTZEendEd1RfSmI0ZkQ0SndBeGgxVTBhTDdpTmstcUZKWG1FQzBK?oc=5","source":"IBM","lang":"zh","date":"2025-10-08"},{"title":"RAG系列：MinerU、Docling还是Unstructured？用OmniDocBench评测开源文档解析工具","url":"https://news.google.com/rss/articles/CBMiU0FVX3lxTFBZSm5QM0ZFaS1PME1MeXNFbVprRWJHcF9BU2RnNXM2WkhVUjZPd09wRFRwczRvUmFmRHhtcGtsdEdmcFVfeGUyZEdScEJEZkhxUWQ0?oc=5","source":"51CTO","lang":"zh","date":"2025-07-16"},{"title":"SmolVLM2: 让视频理解能力触手可及","url":"https://news.google.com/rss/articles/CBMiSEFVX3lxTE52eUJacVFkOEJncTZuSV9jZmlqaGdKZTBJU0NYVjhFRFFRaHpTU0tUemt0NHJVTjRfVGtpQWgyX3RGazVPaEluQw?oc=5","source":"智源社区","lang":"zh","date":"2025-03-22"},{"title":"6大RAG知识库PDF文档处理工具对比！","url":"https://news.google.com/rss/articles/CBMiSEFVX3lxTE94UVczY0dJS2lDS2dBQUtRaVNjM2pDZlM4TS1TcVpXUmNodE44TkpsQW5pTUU2ZlRfNjNmVk5CWmR4U3BxNzc0ZA?oc=5","source":"智源社区","lang":"zh","date":"2025-03-04"},{"title":"微软开源Markdown工具爆了：支持Office文档，可接多模态LLM直出报告","url":"https://news.google.com/rss/articles/CBMiSEFVX3lxTFBUTVk2T1k1TVVoUHI1azJodjJ3NmhKVnRXaHBRU3pxTU1SMmd3azNmUDQ5TnRIZmljWEttRy1jTlBLYVgxcnNZTQ?oc=5","source":"凤凰网","lang":"zh","date":"2025-01-20"}],"stale":false,"snapshot_at":"2026-11-10T09:35:25Z","history_start":"2026-09-06"}