Voice Management
Save voices from VoiceDesign for guaranteed consistency. List and manage your saved voice library.
Workflow
- Generate audio with VoiceDesign
- Save the voice with
POST /v1/voices - Use the voice ID with /v1/audio/speech
/v1/voicesSave a Voice
Extract voice embeddings from VoiceDesign audio and save for future use. Requires API key authentication.
Request Parameters
| Parameter | Type | Description |
|---|---|---|
namerequired | string | Display name for the voice (1-50 characters) |
audiorequired | string | Base64-encoded WAV audio from VoiceDesign generation |
descriptionrequired | string | Original voice description (stored for reference) |
ref_textrequired | string | Transcript of the reference audio. Required for accurate embedding extraction. |
language | string | Language of the voice (e.g., "English", "Spanish", "Japanese") |
# First, generate audio with VoiceDesign and save the WAV file
# Then base64 encode it
AUDIO_B64=$(base64 -i sample.wav)
curl -X POST "https://api.murmr.dev/v1/voices" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d "{
\"name\": \"My Narrator\",
\"description\": \"A warm, professional voice\",
\"audio\": \"$AUDIO_B64\",
\"ref_text\": \"Sample audio text here.\",
\"language\": \"English\"
}"Response
{
"id": "voice_abc123def456",
"name": "My Narrator",
"description": "A warm, professional voice",
"language": "English",
"prompt_size_bytes": 51200,
"created_at": "2025-01-29T12:00:00Z",
"success": true,
"has_audio_preview": true
}/v1/voicesList Saved Voices
Retrieve all voices saved to your account. Requires API key authentication.
curl "https://api.murmr.dev/v1/voices" \
-H "Authorization: Bearer YOUR_API_KEY"Response
{
"voices": [
{
"id": "voice_abc123def456",
"name": "My Narrator",
"description": "A warm, professional voice",
"language": "English",
"language_name": "English",
"audio_preview_url": "https://...",
"created_at": "2025-01-29T12:00:00Z"
},
{
"id": "voice_xyz789ghi012",
"name": "Customer Support",
"description": "Friendly and helpful",
"language": "English",
"language_name": "English",
"audio_preview_url": null,
"created_at": "2025-01-28T10:30:00Z"
}
],
"saved_count": 2,
"saved_limit": 10,
"total": 2
}The saved_limit varies by plan. See limits below.
/v1/voices/:idDelete a Voice
Permanently delete a saved voice from your account. Requires API key authentication.
curl -X DELETE "https://api.murmr.dev/v1/voices/voice_abc123def456" \
-H "Authorization: Bearer YOUR_API_KEY"Response
{
"success": true,
"id": "voice_abc123def456",
"message": "Voice \"My Narrator\" deleted"
}/v1/voices/extract-embeddingsExtract Embeddings
Building a multi-tenant app?
Use Portable Voice Embeddings to store voice data in your own database and pass it via voice_clone_prompt — no saved voice IDs needed.
Extract voice embeddings from audio. This endpoint goes through the Worker at api.murmr.dev and uses API key auth.
| Parameter | Type | Description |
|---|---|---|
audiorequired | string | Base64-encoded WAV audio to extract embeddings from |
ref_textrequired | string | Transcript of the reference audio. Required for accurate embedding extraction. |
model | string | Model to use: "base" (default) or "base_06b" |
curl -X POST "https://api.murmr.dev/v1/voices/extract-embeddings" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"audio": "<base64-wav>", "ref_text": "Transcript of the audio."}'Voice ID Format
Voice IDs follow the pattern voice_ followed by a unique identifier:
voice_abc123def456
voice_xyz789ghi012Voice IDs are case-sensitive and must be used exactly as returned from the API.
Saved Voice Limits
The number of voices you can save depends on your plan:
| Plan | Saved Voices |
|---|---|
| Free | 3 |
| Starter | 10 |
| Pro | 25 |
| Realtime | 50 |
| Scale | 100 |
Error Responses
Bad Request
Missing name or audio, invalid base64 encoding
Not Found
Voice ID doesn't exist or doesn't belong to your account
Conflict
Saved voice limit reached for your plan
See Also
- VoiceDesign API — Create voices from descriptions
- Saved Voices API — Generate audio with saved voices
- Pricing & Usage — Plan limits and quotas