VoiceChangerAPI
Production-oriented speech APIs with credit-based metering, durable jobs, and strict privacy for your users’ data.
Explore every endpoint and run real requests in the browser
Served by Scalar from the live OpenAPI schema (mirrored at /api/openapi). Authenticate like the dashboard (Bearer Firebase ID token) or with X-API-Key from Account → API keys. HTML /admin routes are for operators — lock them behind your edge or VPN.
Built like infrastructure, documented like a product
VoiceChangerAPI is the same stack that powers Voicechanger.co: durable jobs in Postgres, Redis-backed rate limits, presigned object storage, and GPU inference behind a thin FastAPI surface. You integrate once; we absorb spikes, retries, and long-running work.
Whether you are shipping a SaaS feature or an internal tool, you get the same contract our dashboard uses: Firebase Bearer tokens for interactive sessions, or API keys for servers and CI. Credits are reserved when work is queued so usage stays predictable.
Every guide below walks through the story first—who it is for, how the flow feels—then the exact REST calls. Prefer buttons to bytes? The dashboard panels run the identical endpoints with guided UX.
Use the Scalar playground on every page for exhaustive schemas; these guides focus on narrative, correct sequencing, and where to click when you do not want to touch curl.
Choose your integration path
Deep dives with diagrams, accurate REST flows, and a direct link into the matching dashboard workspace.
What runs under the hood
- Postgres job ledger with explicit statuses and credit reservations
- Redis queues + rate limits to protect inference
- S3-compatible storage for uploads, references, and results
- NVIDIA Triton for Omnivoice-class generation and transcription
Ship faster with the dashboard, scale with the API
Non-developers can generate audio, run batch STT, manage dubbing uploads, and mint API keys without reading OpenAPI. Developers get the same endpoints with full automation.
Run the full stack locally with Docker: see voice-changer-api/ops/up.sh