Inference APIs

Text-to-speech playground

Type text, pick a voice, listen — then copy the request into your code.

API docs
Sign in to generate speech
Free account · starter credit included · takes a few seconds
0 characters

About this playground

Speech is generated by the selected model through /v1/audio/speech, the OpenAI-compatible endpoint, and returned as mp3 or wav. Billing is per character of input text. Kokoro returns a paragraph in under a second; Orpheus is slower but more expressive and better suited to pre-rendered narration.

Tips

  • Voice ids encode language and gender: af_ American female, bm_ British male, jf_ Japanese female, and so on. Per-language voice lists are on the languages pages.
  • Numbers and abbreviations are read as written; spell out anything that must be pronounced a specific way.
  • The same request works from the OpenAI SDKs by setting base_url — see the Text-to-Speech API page.