Deploy

Deploy with Code (API & SDK)

View as Markdown

Integrate Voice Agents programmatically with APIs and SDKs. For platform speech APIs, see the API Reference.

What will be covered

TopicDescription
API recipesReady-to-use code for common operations, creating agents, starting sessions, triggering outbound calls, fetching transcripts
WebSocket integrationReal-time voice sessions with audio frames, turn events, and tool-call signals
SDK usageTyped client libraries for session lifecycle, agent management, and outcome retrieval
AuthenticationAPI key usage and best practices for keeping secrets server-side

Authenticate all requests with your API key.

Getting started (preview)

Use the dashboard API recipes as a starting point:

  1. Create or update an agent configuration: set up agents programmatically instead of through the UI
  2. Start a conversation session: initiate a voice or text session via API
  3. Trigger outbound calls: kick off calls to specific numbers with variable payloads
  4. Fetch transcripts and outcomes: pull conversation data into your systems for analysis or CRM updates

WebSocket integration

For real-time voice sessions, connect over the Conversations WebSocket interface. Handle audio frames, turn events, and tool-call signals according to the API contract.

Never expose production API keys in browser code. Keep secrets server-side and proxy WebSocket connections through your backend when building client-facing applications.

Embed the agent (widget)

To put a voice or chat agent on your website, embed the widget:

  1. Open your agent and go to the Embed option in Deploy with Code.
  2. Copy the embed snippet from the dashboard.
  3. Paste it into your site, typically before </body>.
  4. Load the page and confirm the widget initializes.

Customize the launcher position, theme, and which agent the widget uses from the dashboard, and re-test in test agent after changes.