Respeecher is a voice-cloning and speech-synthesis company from the film and games industry that now offers a real-time text-to-speech API, branded Respeecher Space, for voice agents. The API streams audio over a WebSocket with voice streaming under 200 ms, and it is billed pay-as-you-go per 60,000 characters (about one hour of speech) with no subscription.
Alongside the real-time API, a REST gateway exposes the studio workflow: create a project, upload or generate a recording, optionally calibrate pitch, choose a target voice, place a conversion order, and download the result. That gateway covers both speech-to-speech conversion and text-to-speech, with voices filterable by gender, age group, pitch range, nationality, and species. A voice marketplace and a Pro Tools plugin serve production teams.
Features
- Real-time TTS streaming: WebSocket audio streaming with sub-200 ms voice streaming for agent responses
- Voice library: more than 100 male, female, and kids' voices across accents and tones, filterable through the voices endpoint
- Speech-to-speech: convert an existing recording into another voice, with semitone pitch correction
- Voice cloning: enterprise voice cloning delivered as a service by the studio team
- Audio formats: WAV, OGG, MP3, and FLAC on the REST gateway
- Narration styles and accents: per-voice styles for TTS and accents for speech-to-speech
- Rate limits: default of 500 fast and 100 slow requests per 300 seconds per user
- Playground and API keys: API keys are issued from the playground, where you can try synthesis before integrating
- Pricing: $2 per 60,000 characters, pay as you go
