Chat With Your Own Documents: RAG With Source Citations, Explained

Pasting a long document into a chat and hoping the model reads it all is unreliable — and it forgets the moment the conversation moves on. The better pattern is RAG (retrieval-augmented generation): the AI stores your documents, finds the relevant passages when you ask, and answers from your sources — with citations.

one-chat.app — it builds things
Turn that into a budget I can play with.
Done — live budget on the right. Drag any number.
Japan · Aprillive
Flights$1,840
Hotels$1,310
Food + fun$900
Total $4,050 · under budget 

How RAG actually works

1. You upload documents — PDFs, notes, reports.
2. They're chunked and embedded — split into passages and indexed so they're searchable by meaning, not just keywords.
3. On each question, it retrieves — the few passages most relevant to what you asked.
4. It answers from those passages — and tells you which document each fact came from.

The result: answers grounded in your material, not the model's training data — and a citation so you can verify.

Why citations matter

An AI answer you can't check is a guess with confidence. When the reply says "from Q3-report.pdf," you can click through and confirm. That's the difference between a plausible answer and a trustworthy one — especially for work.

How to do it in 1Chat

1. Open a Space and go to Knowledge.
2. Add your files. They're indexed in the background — you'll see each source's status.
3. Just ask. When your question is answerable from your documents, the reply pulls the relevant passages and shows which document it used.

Because Knowledge lives in a Space alongside your memory and connected tools, the assistant can combine your documents with what it already knows about you and the apps it can act in — not just answer from one file in isolation.

Stop pasting walls of text and hoping. Upload once, and ask your documents anything — with the receipts.