Your notes. Your PDFs. In every answer.
You already keep your own summaries, local pathways, and practice protocols. Upload them once and the assistant reads them before replying — quoting the passage that matters and naming the file it came from.
Text files on every plan, PDFs on Pro and Max. Searching your library in chat costs nothing extra.
Upload in Settings.
Go to Settings → Documents and add .txt, .md, or .pdf files. Give each a short, recognisable name — that is the name the assistant will cite.
Ask as normal.
Start an educational or reference conversation in chat. You do not need to mention your documents or tell the assistant to look — it checks the library on its own when the question fits.
Read the citation.
When the reply uses one of your files, the filename appears inline beside the passage. Open it from Settings → Documents if you want the full context.
The best learning answer is often sitting in a document you already wrote.
Best Use Cases
Where it works well Use it
- ✓ Local pathways
Upload CCG or ICB pathways. The assistant reaches for them before generic guidance.
- ✓ Practice protocols
Keep your safety-netting wording or admin protocols on hand and cited back to you.
- ✓ Personal notes
Upload revision notes or case summaries and have them resurface when the topic comes up.
Practical Limits
- ! File formats
Supported: .txt, .md (all plans), .pdf (Pro and Max only). Size caps: 1 MB per text file, 3 MB per PDF.
- ! Document limits
5 on Free, 20 on Pro, 50 on Max.
- ! Readable text required
Scanned or image-only PDFs without a text layer cannot be read — export a text version first.
Generic guidance is fine until it contradicts your local pathway. Your own documents should be part of the educational context when the assistant speaks.
What it costs.
Flat one-time cost to process.
Flat one-time cost to process.
Searching the library is included in standard chat reply costs.
Bring your own knowledge base.
Let the assistant use your trusted documents in its answers.