Material Preparation and Semantic Search
Before writing, users can search historical documents for needed content and rewrite based on the retrieved material, which is suitable for organizing existing materials before drafting.
讯飞文书 is an AI materials writing platform for drafting, reviewing, and recording, helping users improve efficiency in reports, speeches, and minutes.
Xunfei Wenshu is an AI materials writing platform for document-writing users, customized and trained based on the Spark large model and document data. It breaks common materials-writing workflows into stages such as material preparation, draft writing, and review/proofreading, helping users complete a continuous workflow from finding materials to final draft checking within one platform.
From the page display, this product mainly serves transactional and materials-based writing scenarios. It supports generating full text from required elements as well as polishing, expanding, continuing, and imitating existing text. It also provides a voice note-taking feature, which is suitable for meetings and office minutes that need to be recorded first and organized later.
Before writing, users can search historical documents for needed content and rewrite based on the retrieved material, which is suitable for organizing existing materials before drafting.
Generate a full draft with one click based on the filled-in elements, assisted by the large model, covering the drafting stage of materials writing.
After writing is completed, check for typos, awkward wording, sensitive words, and adjust formatting for the review and proofreading stage.
Supports recording while taking notes, and summarizes content from meetings and office scenarios to make post-meeting organization easier.
You can upload a sample article for reference, and the large model will automatically extract elements and imitate the writing, suitable for secondary creation based on existing drafts.
The page also shows an “content processing” direction with polishing, expansion, and continuation capabilities, indicating the product not only generates new drafts but also processes existing text.
When preparing materials for party committees, themed education, or Party building, first use semantic search to find related content from historical documents, then rewrite and integrate it.
Quickly generate a first draft from elements such as a given title, background, work summary, or issue reflection, for drafting organizational reports, speeches, or reflections.
After completing a first draft, check for typos, awkward wording, and sensitive words, then standardize the layout, which is suitable for the proofreading stage before formal submission.
Record and take notes during meetings, research visits, or office communication, then extract key points later to reduce manual organization work.
Upload a reference sample article and let the system extract elements and imitate the style according to the article structure, suitable for scenarios where consistent structure and tone are needed.
Xunfei Wenshu is designed for document writing users, supporting workflow assistance from material preparation and draft writing to review and proofreading, and also providing voice note-taking and draft-based writing features. The page does not clearly restrict specific occupations, but the content is clearly oriented toward materials writing, government affairs, and document scenarios.
Based on the page content, the product covers material preparation before writing, full-text generation during writing, and review and proofreading after writing. It also provides voice note-taking for recording and organizing meetings or office scenarios.
The page clearly shows a free trial, and also provides team edition consultation and Android client download entry points. However, the public information does not provide specific plan names, prices, or trial limitations.
The homepage lists an Android client download entry and provides a team edition consultation entry. The page text does not further indicate whether there are clients for other platforms.
The page explicitly states that compliant content generated on this website will be kept strictly confidential, and mentions that the generative AI service capability is powered by the Spark Cognitive Large Model. Beyond that, the page does not elaborate on more security, permission, or compliance details.