For developers
Already have a UI? Skip the widget. Whitford gives you a single API that answers questions from the documents you choose, with citations, so you can put grounded AI inside your own product without building retrieval yourself.
One request
Authenticate with a secret key (server-to-server, never exposed in a browser). The response includes the answer and the source documents it came from.
Streaming is available at /public/chat/stream (SSE) for token-by-token output.
curl https://api.whitford.app/public/chat \
-H "Authorization: Bearer wf_sk_your_secret_key" \
-H "Content-Type: application/json" \
-d '{"message": "What are your refund terms?"}'{
"text": "Refunds are available within 30 days of purchase...",
"sources": [
{ "fileName": "Returns Policy.pdf", "pageStart": 2, "pageEnd": 2 }
],
"sessionId": "b3f1..."
}Upload the material the API should answer from. Whitford builds and indexes an AI Brain from it.
Generate a server-side API token in your dashboard. It’s scoped to your firm and rejected from browsers.
POST a message to /public/chat from your backend and render the grounded answer + citations in your own UI.
Pricing
The API is included in Customer Chat ($49/mo). It’s the same grounded engine as the website widget, just headless. Usage runs against the same monthly allowance, and you can top up any time if you need more.
Live in minutes. No retrieval stack to build.