scid.aiCollect your own material
Upload documents and tables that did not come from scientific search.
Your files remain part of the research
Keep uploaded material and documents created in scid.ai in one Library. Add them to a new or existing research session so document content and spreadsheet rows become part of the answer context.
A file system inside the research workspace
PDFs, tables, text files, and exported documents do not have to disappear into computer folders. They remain close to the research and become sources again when needed.
Remember where a file came from
The Library shows the file type and its related chat when the material was created during research.
Start research from files
Select the documents you need and add them to the current context or open a new chat.
One list for incoming and created material
The Library brings together your uploads and documents created in scid.ai. Search, filters, and sorting make it easy to return the right material to the work.
- Working formats are supportedPDF, TXT, Markdown, DOC/DOCX, CSV/TSV, JSON, XML, HTML, RTF, TeX, and Excel workbooks.
- Files become research sourcesDocument content and real spreadsheet rows can be used in research chat.
- Take the material back outDownload one file or several selected files as a single ZIP archive.
The Library works with the research
Files stay connected to the question, the sources, and the output.
Upload and find
Add your material, search by name and related chat, and filter by type or format.
Use it as evidence
Select files and attach them to the current question or start a separate research session.
Keep generated output
Documents exported from chat as PDF, DOCX, or TeX return to the Library.
Questions about the Library
Can I begin research with my own files?
Yes. Select Library files and start a new chat with them or add them to the current sources.
Are documents created in scid.ai retained?
Yes. PDF, DOCX, and TeX files exported from chat can be saved in the Library as generated files.
How do I find a file?
Search by file name or related chat and use filters for format and source type.
Make your material part of reusable research memory
Upload documents once and bring them back into context whenever the question depends on those data.