how it works
From folder to assistant in three steps.
01
Upload
Drop in PDFs, DOCX, TXT, or Markdown. Docubix chunks, embeds, and indexes everything automatically.
02
Configure
Write a system prompt, pick a model, and tune retrieval — no ML experience required.
03
Integrate
Call the REST API from your app or backend. Your assistant is live.
product tour
See Docubix in two minutes.
Upload documents, ask questions, and get cited answers — a quick walkthrough of the full workflow.
try it
Ask it something.
A live preview of Docubix chat. Pick a knowledge base, then try one of the questions below.
QA customer can't log in — what should I tell them?
Have them tap Forgot password on the login screen. They'll get a one-time reset link by email within a few minutes. Links expire after one hour — if nothing shows up, ask them to check spam or confirm the email on the account.
features
Everything a RAG product needs. Nothing it doesn't.
Knowledge bases
Spin up a separate assistant, vector index, and API key for every project you're working on.
Document uploads
PDF, DOCX, TXT, and Markdown. Drag a file in and Docubix handles chunking and embeddings.
Citations
Every response links back to the exact document and page or section it came from.
Streaming responses
Tokens stream as they're generated, so answers feel instant — not like a form submission.
API access
One REST endpoint for chat, history, and search. Build whatever interface you want on top.
Analytics
See what people are asking, which documents get cited most, and where your docs fall short.
for developers
Three lines, one assistant.
No ML experience. No vector DB setup. One endpoint handles retrieval, generation, and citations — stream the response or wait for the full payload.
works with
const res = await fetch("https://api.docubix.com/api/v1/chat", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
api_key: "pk_live_...",
message: "How do I cancel my subscription?",
}),
});
const { answer, citations } = await res.json();use cases
Built for the jobs founders actually have.
Add AI chat to your product
Stop building a support bot from scratch — wire ours in with one API call. Your users get cited answers; you skip the RAG pipeline.
api/chat.tsCustomer support
Deflect repetitive tickets by answering from your help center automatically, citing the article used.
help-center/billing.mdInternal company AI
Turn your handbook, wikis, and SOPs into an assistant your whole team can query.
handbook.pdfDocumentation assistant
Replace "search the docs" with "just ask". Answers link straight to the page that explains it.
docs/getting-started.mdpricing
Simple pricing. No token math.
Two plans. Upgrade when you outgrow the free tier — not when you hit a confusing usage cap.
Free
For individuals exploring document AI
$0/ month
- 1 knowledge base
- Up to 20 documents
- 100 queries / month
- 0.25 GB storage
- API access
- Citations on every answer
Pro
For teams shipping AI to production
$79/ month
- 10 knowledge bases
- Up to 1,000 documents
- 3,000 queries / month
- 10 GB storage
- API access
- Citations on every answer
faq
Questions, answered.
01What file types can I upload?↓
PDF, DOCX, TXT, and Markdown today. HTML and Notion sync are next.
02Do answers always include citations?↓
Yes. Every response includes at least one source with a document name and page or section.
03Can I use the API directly?↓
Yes — every plan includes API access. Pro adds unlimited keys for separating environments.
04Is chat history saved?↓
Conversations are saved per user, so people can pick up where they left off — and you can review them later.
05How many knowledge bases can I create?↓
One on the Free plan, unlimited on Pro. Each has its own documents, settings, and API key.
06Do I need to train or fine-tune a model?↓
No. Docubix handles chunking, embeddings, and retrieval — you bring the documents.
get started
Ship cited AI chat this week — not next quarter.
Free plan, no credit card. Upload docs, hit the API, and skip building embeddings infrastructure yourself.