Chat with up to 4,000 files

The scienceOS paper index is based on more than 240 million research papers from the Semantic Scholar index, along with 1.0 billion full-text snippets and 2.9 billion reference pairs. When you ask a question, the answer is grounded in information like the full text of open access papers, abstracts from closed access papers, and other relevant data from these papers.

The scienceOS paper and citations database is updated on a weekly basis.

We are committed to strict EU data‑residency and processing controls, and provide a fully GDPR-compliant AI research agent. All our AI and machine learning models, as well as your chats, your uploaded files, and your user data are located on servers within the European Union. For holders of Team or Institution plans, we provide a Data Processing Agreement (DPA) upon request, as well as support for meeting the compliance requirements of the EU AI Act.

Neither user-generated inputs (e.g. your prompts or uploaded files) nor user-generated outputs (e.g. text answers or image files) are used to train new scienceOS AI models.

When you upload a file to your scienceOS account, it is stored securely, automatically encrypted, and accessible only to you. If you create a project and add your files to it, you can choose to share access with colleagues and others. People you invite to a shared project can read and interact with your files and may save the file’s metadata (title, abstract, authors, etc.) to their own library. You can also suggest files from your library to be added to shared projects you have been invited to, and choose whether to allow the attached files to be copied and saved to the project owner’s account.

You can upload any .pdf, .docx, .doc, .odt, .rtf, .pptx, .ppt, and .odp file with a maximum file size of 80 MB. On Free and Angel plans, up to 128 pages per document are processed; on Team and Institution plans the limit is 1024 pages per document. During the upload, all documents are transformed into PDFs. Free users can extract text that is already encoded in the transformed PDFs. For users on paid plans, documents are enriched to also extract text from scanned documents as well as figures and tables. If a file has not yet been enriched, re-uploading it allows users on paid plans to trigger enrichment.

Yes, you can use scienceOS for free and access our fast AI model. To access our best model and all other features, consider subscribing to our Angel plan.

You can subscribe to the scienceOS Angel plan for access to all features of the scienceOS AI research agent. This includes: unlimited use of our fast model, extracting and chatting with figures from your PDFs, and access to our best model. Funds from Angel plan subscribers are only used to cover costs, develop new tools, and sustain the free plan.

When using scienceOS, it is important to uphold academic integrity. Many journals require authors to disclose AI tools in the methods section or equivalent. We recommend specifying how you used scienceOS – whether for editing, translation, or research – and including the version and link in your description (e.g., scienceOS (June 2026 version), https://www.scienceOS.ai). Including a transcript of your interactions as an appendix can further demonstrate your commitment to transparency.

Download the Use of AI disclosure template or review an example for how other scientists reference scienceOS in their papers.

We believe in collaboration. If you, your institution, organization, or company would like to work together with scienceOS, please use [email protected] and connect with us. We are excited to discuss how we can work with you.