Streaming vs batch
The API transcribes finished files (batch) and live audio over WebSocket (streaming).
Batch (REST)
A single POST to /api/v1/transcribe. You hand the server the whole clip, the server hands you back the whole transcript when it's done. Latency grows with the length of the clip: the transcript arrives once the whole file has been processed.
Use when: you have a finished recording on disk, you don't need partial results, or your client lives behind a strict firewall.
Streaming (WebSocket)
A persistent connection on /api/v1/stream where you send live audio and get text back while the person is still speaking. Open it from your server: the API key goes in the first message.
Use when: live captions, real-time interpretation, voice agents, long sessions you can't fit in a single upload.
Which to pick
Today: batch, for everything. When streaming launches, the rule of thumb will be: if a person is waiting for text in real time, stream. If a pipeline processes a file, use batch. Both are billed the same way, per second of audio.