Space APIIntegrations

Ultravox Integration

The SSE endpoint is compatible with Ultravox’s bring your own TTS feature.

This integration imports a voice that is already available to your account. It does not clone a voice—see Obtain a list of available voices.

1

Import a voice

Go to https://app.ultravox.ai/voices and select Import voice. Give it a name and, optionally, a description and a primary language.

Use Import voice, not Create custom. The latter opens Ultravox’s voice cloning panel, which defaults to ElevenLabs and cannot be used with Respeecher TTS.

2

Configure with JSON

Paste the following JSON, substituting <API_KEY> with your Respeecher API key:

{
    "url": "https://api.respeecher.com/v1/public/tts/en-rt/tts/sse",
    "headers": {
        "X-Api-Key": "<API_KEY>",
        "Content-Type": "application/json"
    },
    "body": {
        "transcript": "{text}",
        "voice": {
            "id": "samantha"
        },
        "output_format": {
            "encoding": "pcm_s16le"
        }
    },
    "responseSampleRate": 22050,
    "jsonAudioFieldPath": "data",
    "responseMimeType": "application/jsonl"
}

Test your voice with the respective button and, if it works, save it. You can change the en-rt endpoint to one of our other models as well as change samantha to another voice ID from the voices available to your account. The voice object can also contain sampling_params.