Integration

Build on the SupportAi REST API

The widget covers most teams, but some need the agent inside their own product, mobile app or internal tool. The REST API exposes everything the widget does.

Stream replies token by token, receive tool-call events as actions run, manage knowledge sources programmatically and react to conversations, leads and escalations through signed webhooks.

Last updated

Setup

  1. 1

    Create a workspace API key with the scopes you need

  2. 2

    Send a message to /chatbots/{id}/chat with JSON or SSE

  3. 3

    List and manage sources, conversations and leads

  4. 4

    Subscribe to webhooks for real-time events

Read the docs

What you get

  • Chat endpoint with SSE streaming and tool-call events
  • Conversation and lead read endpoints
  • Knowledge source create, update, delete and reindex
  • Scoped keys: read, write, chat
  • OpenAPI 3 spec and downloadable docs

Chat with streaming

POST a message and receive JSON, or request SSE for streaming tokens, widgets and tool events. Conversation state is handled server-side; pass a conversation ID to continue.

Automate knowledge

Create sources from URLs, upload files, add Q&A pairs and trigger reindexing from your CI or CMS so the agent is always current.

Security and limits

Workspace-scoped keys with read, write and chat scopes, rate limits per key, hashed key storage and HMAC-signed webhooks with retries and replay.

FAQ

Frequently asked questions

The API is plain REST with an OpenAPI 3 spec; generate a client in any language or use fetch.

Keep exploring

Read the API docs

OpenAPI spec, examples and webhooks reference.

Start free Book a demo

No credit card · Live in minutes · Cancel anytime