Choosing an API
Respeecher offers two separate APIs. They have different base URLs, different authentication, and different strengths — so it’s worth picking deliberately before you start building.
Realtime, low-latency text-to-speech streaming. The right default for most TTS use cases, including voice agents.
Speech-to-speech conversion and project-based workflows: projects, folders, recordings, calibration, and conversion orders.
At a glance
The two APIs use different API key headers — X-API-Key for Space and
api-key for Marketplace — and the keys are issued from different places. Keys
are not interchangeable between the two.
Choose the Space API if…
- You are generating speech from text and want it as fast as possible.
- You need to stream audio as it is generated, over SSE or WebSockets.
- You are building a voice agent or any interactive experience where latency matters.
- You want a Python or TypeScript SDK, or a drop-in integration with LiveKit, Pipecat, Ultravox, or VAPI.
Start with the Quickstart.
Choose the Marketplace API if…
- You are converting existing audio from one voice to another (speech-to-speech).
- You want to organize work into projects, folders, and recordings rather than one-off requests.
- You need calibration to match pitch between the source and target voice.
- You want to convert to a specific accent, or export a whole project as a
.zip. - You are working in Avid Pro Tools and want the plugin.
Start with the API Overview.
Still not sure?
If your use case is “turn this text into audio,” use the Space API. It is faster, simpler to authenticate against, and has SDKs.
Reach for the Marketplace API when you specifically need speech-to-speech conversion or the project/folder/recording workflow around it.