Wirestack
OpenAI prompt-to-voice opens to all API customers on October 16, 2026: free custom GPT-Live voices from a text description
Describe a voice in text and POST it to /v1/audio/voices. Creating it is free; up to 100 per org; works only with GPT-Live.
Your keys
0 of 1 ready
Create each key on its own page, paste it below, then copy the finished lines into your project's .env file. What you paste stays in this tab. It is never sent to Wirestack or saved.
- OpenAI API keyOPENAI_API_KEYGet key
Click Create new secret key
What you get
On October 16, 2026 OpenAI made prompt-to-voice available to all API customers. You describe a voice in text and get a reusable custom voice for gpt-live-1 sessions. Creating the voice is free. Speech made with it is billed at regular GPT-Live session rates.
The guide says to send a request to /v1/audio/voices with type set to prompt, plus a name and the prompt description. An optional script_hint can be up to 2,000 characters. Each organization can create up to 100 prompt-generated voices. Voices must be adult voices. They work only with GPT-Live, not the Text to Speech, Realtime or Chat Completions APIs.
The same entry added nine GPT-Live voices: Nuri, Juni and Flitz (German), Brise and Sillage (French), Nira (Hindi), Harema and Shida (Japanese), and Noeul (Korean).
Prerequisites
Steps
- Describe the voice you want and create it through the official guide or the voice Playground.
- Save the returned voice ID.
- Set
session.audio.output.voice.idto that ID when you start a session, and put the description insession.instructions. - To try another voice, start a new session.
Test it
Create one voice, start a GPT-Live session with its ID, and listen to the reply.
Not confirmed
- A maximum length for the voice description.
- Whether prompt-generated voices will come to other OpenAI audio APIs.