Google has added the ability to create and directly edit documents to the Gemini app. This update, announced on October 29, allows users to create files within a dialog box and edit them while previewing the results. It supports documents, PDFs, presentations, spreadsheets, code, and more.

Users can expand the generated file to full screen to directly input content. They can also select text to ask questions about specific sections or request corrections from Gemini. The system is designed to allow users to seamlessly continue both manual editing and instructing the AI within the same document.

The example introduced by Google describes the process of converting handwritten lecture notes into PDF learning materials. After uploading the notes and requesting the creation of the material, users directly edit the content while reviewing the output. It is also possible to select specific sections to request modifications to text or visual elements, and to ask follow-up questions. The workflow encompasses the review and revision process, even after the file is created, all within the conversation.

The interface change focuses on connecting the conversation content with the edit target. When a user selects a section to modify within the document, questions and requests follow based on that location. Instructions for the entire document and feedback on specific parts can be used together. Full-screen editing takes on the role of focusing directly on the document itself for modification.

Supported environments vary by file format. PDF preview and editing are available on the web and mobile. Google Docs, Sheets, Slides, Markdown, text files, and code are supported on the web. Google announced that it is making this feature available to all users and supported languages in the regions where the Gemini app is available.

See more and sources
https://blog.google/innovation-and-ai/products/gemini-app/edit-files-directly/