Oto Voice API

Oto Voice API turns spoken words into structured data and lets you keep every conversation—and its follow-ups—fully accessible and actionable.

What is Oto Voice API

With the Oto Voice API you can:

• Stream live audio via WebSocket and get real-time speech-to-text plus automatic “action” extraction (to-do items, calendar events, research queries).

• Query, update, and track those actions over REST—mark them accepted, completed, or deleted from any client.

• Manage conversations: list past calls, download raw audio, grab full transcripts, or pull quick summary logs for instant context.

How it works

The Oto Voice API has two faces:

1. WebSocket – for real-time audio streaming, transcription, and action detection

2. REST – for browsing and updating actions and conversations

Authentication

Authorization: $OTO_API_KEY
OTO_USER_ID: {user_id}

Typical Workflow

1. Open WebSocket at /conversation/{id}/stream, stream audio frames.

2. Receive live transcribe updates plus detect-action objects.

3. Get or filter with GET /actions?conversation_id={id} to see everything detected.

4. Update an action’s lifecycle (PATCH /action/{id}) as the user confirms, completes, or dismisses it.

5. Complete the conversation to finalize the audio stream.

6. Retrieve the final transcript or download the audio via the Conversation endpoints.