TTS API, OpenAI-Compatible with Zero Migration
Compatible with OpenAI /v1/audio/speech format — copy the code and integrate. Supports streaming, SSML, and multi-speaker dialog.
Developer Pain Points
- ✕Self-hosting TTS is costly and hard to maintain
- ✕Migrating to a new API means rewriting lots of code
- ✕Low voice quality hurts product experience
- ✕Per-minute billing is opaque and hard to control
The OpenAI-Compatible TTS API Solution
OpenAI-compatible — copy the SDK examples and integrate. Transparent per-character billing, 30+ premium voices. Pro users get 100k chars/mo for $29.9.
Integrate the API in 3 Steps
Sign up and create an API Key (sk-wxtts- prefix)
Copy the Python/Node.js/cURL examples
Call /v1/audio/speech to generate audio
FAQ
Is it really OpenAI-compatible?▾
Yes — the /v1/audio/speech endpoint matches OpenAI exactly. Just swap the base_url and API Key to migrate.
How is the API billed?▾
Per-character billing. Pro plan ($29.9/mo) includes 100k chars, overage billed by usage. 60 RPM and 5 RPS limits.
Is streaming synthesis supported?▾
Yes. POST /v1/audio/speech/stream pushes PCM by sentence with <200ms first-byte latency, ideal for real-time narration.
Which plans can use the API?▾
API access is Pro-only. Free and Creator users should use the dubbing tool for voice synthesis.