AI-powered OCR in ONLYOFFICE PDF Editor: how to convert images into text with ease
Version 9.0 brought some key AI improvements aimed at boosting your productivity when working with spreadsheets and PDF files. In this guide, we will focus on the newly added OCR feature and explain how to turn images into editable text using your favourite AI assistant in our PDF editor.
What is OCR?
OCR stands for Optical Character Recognition. It is a technology that enables scanning and converting different types of documents, images and PDF files into editable and searchable data.
In ONLYOFFICE, the OCR feature is backed by artificial intelligence technology, allowing you to turn images into text within seconds using the OCR capabilities of your AI assistant. It works in documents, spreadsheets, presentations and PDF files.
Let’s find out what you need to get started with OCR tasks in your PDFs.
Choose the right ONLYOFFICE solution
First of all, you need to choose the right ONLYOFFICE solution depending on your needs:
- Deploy ONLYOFFICE Docs, the self-hosted version of the suite, on your Windows or Linux server and integrate it with the platform of your choice to create a secure collaborative environment, for example, within Nextcloud, ownCloud, Redmine, Moodle, Alfresco, Jira, SuiteCRM, Seafile, etc. — more than 40 options in total. The cloud-based version of ONLYOFFICE Docs is also available for those who don’t want to install and configure anything manually.
- Create a free ONLYOFFICE DocSpace account and get access to a room-based collaborative platform designed for efficient teamwork in real time.
- Install ONLYOFFICE Desktop Editors on your PC or laptop with Windows, Linux or macOS and enjoy offline document editing for free.
All these solutions feature ONLYOFFICE PDF Editor, a powerful tool that can be used to edit the existing text and add new text boxes, insert and modify visual objects, add interactive fields of different types, leave comments and drawings, manage pages and annotate your documents.
In ONLYOFFICE PDF Editor, you can enable the OCR feature by installing and configuring the AI plugin. Let’s discover how it works in practice.
Enable the AI integration
To run OCR in PDF files, you need to add a suitable AI model via the AI plugin which supports all popular chatbots and tools — ChatGPT, Google Gemini, Claude Sonnet, Mistral, Groq, DeepSeek, xAI, Stability AI and more.
Let’s go through the installation and configuration process in ONLYOFFICE PDF Editor.
Step 1. Install the AI plugin
Open your PDF file and access the ONLYOFFICE plugin marketplace via the Plugin Manager located on the Plugins tab. Find and install the AI plugin by pressing the corresponding button.
Step 2. Enable the AI plugin
Now activate the AI plugin by switching its slider in the Background plugins section located next to the Plugin Manager. A new tab will appear on the top toolbar — AI.
Step 3. Add a suitable AI model with OCR capabilities
Switch to the AI tab and click the Settings icon. You will see the AI configuration menu where you can add and select AI models for different tasks.
Click Edit AI models to add a new model. In the AI Models list menu, you can add and delete your models at any time.
Click the + button and choose one of the providers from the drop-down list in the Name field. In our example, it’s Mistral but you can choose any other option.
After selecting a provider, the corresponding URL parameter is established automatically, and you just need to enter a valid API key in the Key field. You can get one on the official website of your provider.
Please note that some providers don’t provide their API keys for free so you might need to buy one. That’s why it’s highly recommended to consult their policies and terms of use before using their AI models in ONLYOFFICE.
Now choose a model from the Model drop-down list. If you see nothing, click Update models list. Select the model you like but don’t forget that different models work well if you use them for specific tasks. For example, some are good at text generation but are not capable of recognizing text in images. You’d better select something designed for OCR work. You can find more information about the models’ functionality on your provider’s website.
Additionally, you can rename the added model in the Model name field. Please pay attention to the Use model for section — your model should have the Visual Analysis option selected. Otherwise, you won’t be able to use it to run OCR. If your model has OCR functionality but this option is not selected by default, select it manually. To finish, click OK.
Step 4. Assign the OCR tasks
After successfully adding your AI model, you will get back to the AI models list menu. Click the Back button to access the AI configuration window where you will be able to assign your AI model as the default option for OCR tasks. Click OK to finish the configuration process.
Congratulations! Now everything is ready to run OCR in your PDFs. You can access the OCR feature on the AI tab. Is the OCR icon grey? Don’t worry. You just need to switch to the Editing mode because the default Commenting mode doesn’t allow you to recognize text. Either click the Edit PDF button on the AI tab or select the Editing mode in the upper right corner of the interface.
How to convert images into text
Now that you have configured your AI model, you can easily convert images in your PDF files into editable text by clicking the OCR button on the AI tab. The AI assistant will recognize the text on your image and insert it on the same page. Now you can edit it using the vast variety of tools provided by ONLYOFFICE PDF Editor — choose among various fonts and change their size and colour, create bulleted and numbered lists, align text in different ways, inset and modify objects, and more.
Get ONLYOFFICE PDF Editor and use AI for OCR tasks
Use the OCR capabilities of artificial intelligence to turn images into editable text in your PDFs online by creating a free ONLYOFFICE DocSpace account or offline by downloading the free desktop application for your PC or laptop with Windows, Linux or macOS:
Frequently asked questions
Here are the answers to the most popular questions about AI-powered OCR in ONLYOFFICE:
Question: I don’t trust AI tools. Can I run OCR in ONLYOFFICE without using the AI plugin?
Answer: When it comes to PDF files, AI-powered OCR is the only available option. However, if you work with text documents, you can use the OCR plugin which converts images into editable text and doesn’t need AI integration. Read this article to learn more.
Question: Can I run OCR in PDF files only?
Answer: The AI-powered OCR feature is also available in documents, spreadsheets and presentations. To use it, you need to select the required image, access the context menu with a right click and choose AI -> Image -> OCR.
Question: Is any AI model good for OCR purposes?
Answer: Not all AI models can handle such tasks. Before using an AI model for text recognition in ONLYOFFICE, please check its specifications on the provider’s website.
Question: Is AI-powered OCR a paid feature?
Answer: No, the ONLYOFFICE AI plugin and all its features, including OCR, are available for free. However, your AI provider might charge you for using its models in ONLYOFFICE. Please find more information about pricing on your provider’s website.
Create your free ONLYOFFICE account
View, edit and collaborate on docs, sheets, slides, forms, and PDF files online.