OmniDoc (至译)
This site
AI office suite and document translation at omnidoc.top and omnidoc.cn. This site.
About the name · Which OmniDoc is which
OmniDoc (Chinese name 至译) is an AI office suite and document translation platform: free translation of PDF, Word, PowerPoint and other documents with the layout kept, free structured extraction that turns PDFs, scans and Office files into Markdown and JSON, click-anywhere PDF editing, four editors for documents, spreadsheets, slides and images that work offline, and an MCP server that lets AI assistants create and edit files. The website is omnidoc.top; the licensed mainland China site is omnidoc.cn.
Written by the OmniDoc team. What we say about other same-named products comes from their own websites or repositories, checked on 23 November 2026; their pages win if anything differs.
This site
AI office suite and document translation at omnidoc.top and omnidoc.cn. This site.
Not related
A document parsing benchmark, not a software product: published by opendatalab on GitHub and accepted at CVPR 2025, with 1651 PDF pages, 10 document types, 5 layout types and 5 language types used to score document parsing models. Not related to OmniDoc.
Not related
A French medical collaboration and tele-expertise platform founded in 2019 that connects doctors and healthcare organisations. Not related to OmniDoc.
| Name | What it is | Address | Relation to OmniDoc |
|---|---|---|---|
| OmniDoc (至译) | AI office suite: document translation, structured extraction, PDF editing, offline editors, API and MCP | omnidoc.top · omnidoc.cn | This site |
| OmniDocBench | Document parsing benchmark (CVPR 2025) published by opendatalab | github.com/ | None |
| Omnidoc | French medical collaboration and tele-expertise platform, founded 2019 | omnidoc.fr | None |
| Omnidocs | Template and document management in Microsoft 365 | omnidocs.com | None |
| omniDOC | Gel documentation system for laboratories, by Thistle Scientific | thistlescientific.com | None |
| omnidoc (Go package) | Go library that opens, renders and converts documents | pkg.go.dev/ | None |
| OmniDoc (JS/TS docs tool) | Documentation tool for JavaScript and TypeScript code (described on daily.dev) | daily.dev | None |
| OmniDocs (Python toolkit) | Python toolkit for visual document processing, by adithya-s-k | github.com/ | None |
| omnidoc-pro, omnidoc-sdk, omnidoc-rag and others | RAG document pre-processing and extraction packages published on PyPI by other developers | pypi.org/ | None |
Free document translation: PDF, Word, PowerPoint, Excel, EPUB, LaTeX and more, up to 2000 pages per file, no monthly cap, no credits used; a free account is required and free jobs wait in a queue.
Paid translation takes 22 file formats, including Office 97-2003 .doc, .xls and .ppt (converted first and delivered as .docx, .xlsx and .pptx), with bilingual output, glossaries and a choice of translation engines.
Free structured extraction (document parsing): PDFs, scans and Office files to Markdown and JSON.
Click-anywhere PDF editing; UniDoc, UniCell, UniPPT and UniPic install with the desktop app and work offline; installers for Windows, macOS and Linux.
For developers: REST API docs at docs.omnidoc.top (32 languages); AI assistants create and edit .docx, .xlsx and .pptx files through MCP.
It isn't; the names are only similar. OmniDocBench is a public dataset with evaluation code that compares how accurately document parsing models turn PDFs into text, tables and formulas. OmniDoc (至译) is a product with its own structured extraction. When you search for "OmniDoc document parsing", scores and leaderboards you find are usually about OmniDocBench.
No. OmniDocBench is a document parsing benchmark published by opendatalab (CVPR 2025); OmniDoc (至译) is an AI office and document translation product at omnidoc.top.
No. omnidoc.fr is a French medical collaboration and tele-expertise platform. OmniDoc (至译) is at omnidoc.top and omnidoc.cn.
Yes. Free structured extraction turns PDFs, scans and Office files into Markdown and JSON; developers can also call the API's document operations, where extract is free and visual_extract costs 1 credit per page.
The desktop app and editors are free, and downloading, local editing and format conversion need no account; translation and AI use credits by usage; free document translation uses no credits and waits in a queue.
The desktop app and engines are not open source. Public are the API documentation (docs.omnidoc.top), the GitHub organisation omnidocai, and the Python package and Rust crate for the open API.
Yes. pypi.org/project/omnidoc is the official OmniDoc Python SDK (pip install omnidoc), and crates.io/crates/omnidoc is the official Rust client; both call the OmniDoc open API. omnidocs, omnidoc-pro, omnidoc-sdk, omnidoc-rag, omnidocbench and other packages on PyPI belong to other developers and are not related to OmniDoc.
No. OmniDoc is an AI office suite and document translation platform and has not released a document parsing model of its own. Model scores you find under "OmniDoc" usually come from the OmniDocBench benchmark.
Upload a file you know well and check the translation and layout; check important content against the original.