Speech-to-Text

Transcribe audio to text. Multipart upload; cpod-backend forwards to the governed gateway which routes to a multimodal model (e.g. google/gemini-2.5-flash).

• POST/api/v1/stt/transcriptionsTranscribe audio (multipart: file + language + model + metadata)

  Run

Calls go through the portal proxy with your admin session. SSE (run / generate) endpoints return their final payload here; live streaming lands next.