OpenAPI · Live-Playground

Alle Endpunkte erkunden und echte Requests im Browser ausführen

Scalar lädt das live OpenAPI-Schema (gespiegelt unter /api/openapi). Authentifizierung wie im Dashboard (Bearer Firebase-ID-Token) oder mit X-API-Key aus Konto → API-Schlüssel. HTML-/admin-Routen sind für Betrieb — hinter Edge oder VPN absichern.

Roadmap

Realtime STT: streaming is coming, batch is live

Low-latency captions belong on a WebSocket—not a giant POST. We are carving that path, but production workloads already ship with our batch transcription API.

Streaming microphone path (preview)

Be honest about timelines

The dashboard’s Realtime speech to text page is a microphone readiness checklist today. It exists so you can validate hardware permissions while you design your streaming client.

Behind the scenes we are aligning the same credits, auth, and observability you already trust from batch jobs—so when WebSocket ASR lands, you will not need a second billing model.

Until then, batch STT already covers podcasts, meetings, compliance reviews, and any workflow where a few seconds of latency is acceptable.

What to do in production now

Treat this page as planning guidance. Executable REST lives under the batch speech-to-text guide.

  1. 1

    Prototype UX with the dashboard

    Use the realtime page to confirm browser capture and UX copy. Pair it with the batch guide to move audio into the live pipeline.

  2. 2

    Upload with presigned URLs

    Chunk or finalize recordings, then follow POST /v1/stt/uploads → PUT → POST .../start exactly as documented for batch STT.

  3. 3

    Poll jobs like any other workload

    GET /v1/jobs/{job_id} gives you the same statuses as TTS or dubbing, which keeps your observability story unified.

  4. 4

    Watch release notes

    When streaming endpoints ship, they will appear in Scalar first—same auth headers, tighter latency budgets, and the same credit ledger.

Batch fallback (reference)

This is identical to the batch STT flow—your near-term production path.

POST /v1/stt/uploads?content_type=audio/wav
PUT $upload_url
POST /v1/stt/jobs/{job_id}/start
GET /v1/stt/jobs/{job_id}/transcript
Prepare your realtime UX
Use the dashboard page to rehearse microphone flows with your team while backend engineers wire batch transcription.

Need streaming today? Pair batch STT with client-side chunking until the WebSocket service graduates from roadmap to GA.