Providers
OpenAI
OpenAI text-to-speech through the Speechbase gateway — gpt-4o-mini-tts, tts-1, and tts-1-hd.
| Prefix | openai |
| Default model | gpt-4o-mini-tts |
| Provider key | Connect under Provider Keys |
Route to OpenAI by prefixing the model with openai/. Omit the model and pass
openai alone to use the default, gpt-4o-mini-tts.
Models
| Model | Streaming | Audio tags | Timestamps | Max input |
|---|---|---|---|---|
gpt-4o-mini-tts | Yes | Yes | Gateway-generated | 4096 |
tts-1 | Yes | — | Gateway-generated | 4096 |
tts-1-hd | Yes | — | Gateway-generated | 4096 |
Usage
curl -X POST https://api.speechbase.ai/v1/audio/speech \
-H "Authorization: Bearer $SPEECHBASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"mode": "inline",
"model": "openai/gpt-4o-mini-tts",
"voice": "alloy",
"text": "Hello from Speechbase!",
"output": "mp3"
}' --output hello.mp3Built-in voices: alloy, ash, ballad, coral, echo, fable, nova,
onyx, sage, shimmer, verse. The older tts-1 and tts-1-hd models
don't support ballad or verse.
Steering delivery
gpt-4o-mini-tts accepts an instructions provider option that shapes tone,
pacing, and emotion:
{
"mode": "inline",
"model": "openai/gpt-4o-mini-tts",
"voice": "coral",
"text": "Your order has shipped.",
"providerOptions": {
"instructions": "Speak in a warm, upbeat customer-support tone."
}
}Provider options
Anything in providerOptions is forwarded to the OpenAI speech API unchanged,
using OpenAI's own field names (for example speed or response_format).