Documentation

Audio generation

Text-to-speech, voice conversion, isolation, dubbing, and sound effects.

Create and transform audio for voiceovers, localization, podcasts, or sound design. All endpoints live under https://api.thefluxtrain.com/api/v1/audio-generate/ and return the same job record shape.

Authentication: x-api-key on every request

Optional fields (all audio endpoints):

Field	Type	Default	Description
`folder_id`	string	—	Folder UUID in your account
`project_id`	string	—	Optional project ID
`screen`	string	`studio`	Label for your analytics (optional)
`source`	string	`flux-toolkit`	Label for your analytics (optional)
`message_id`	string	—	Your own tracking ID

Text to speech

URL: POST https://api.thefluxtrain.com/api/v1/audio-generate/text-to-speech

Field	Type	Required	Default	Description
`text`	string	Yes	—	Text to synthesize
`voice_id`	string	Yes	—	ElevenLabs voice ID
`model_id`	string	No	`eleven_multilingual_v2`	TTS model
`output_format`	string	No	`mp3_44100_128`	ElevenLabs output format

{
  "text": "Welcome to TheFluxTrain.",
  "voice_id": "your-voice-id"
}

Speech to speech

URL: POST https://api.thefluxtrain.com/api/v1/audio-generate/speech-to-speech

Field	Type	Required	Default	Description
`audio_url`	string	Yes	—	Source audio URL
`voice_id`	string	Yes	—	Target voice ID
`model_id`	string	No	`eleven_multilingual_sts_v2`	STS model
`output_format`	string	No	`mp3_44100_128`	Output format
`filename`	string	No	—	Optional filename hint

Audio isolation

URL: POST https://api.thefluxtrain.com/api/v1/audio-generate/isolation

Field	Type	Required	Description
`audio_url`	string	Yes	Source audio URL
`filename`	string	No	Optional filename hint

Dubbing

URL: POST https://api.thefluxtrain.com/api/v1/audio-generate/dubbing/start

Field	Type	Required	Description
`audio_url`	string	Yes	Source audio URL
`target_lang`	string	Yes	Target language code (2–10 chars)
`filename`	string	No	Optional filename hint

Sound effects

URL: POST https://api.thefluxtrain.com/api/v1/audio-generate/sound-effects

Field	Type	Required	Default	Description
`text`	string	Yes	—	Sound effect prompt
`duration_seconds`	number	No	—	0.5–30 seconds
`prompt_influence`	number	No	`0.3`	0.0–1.0
`loop`	boolean	No	`false`	Seamless loop

{
  "text": "Soft rain on window",
  "duration_seconds": 3
}

Response (all endpoints)

Field	Type	Description
`id`	string	Job ID
`url`	string	Output audio URL
`status`	string	`SUCCESS`, `PENDING`, etc.
`media_type`	string	`AUDIO`
`meta`	object	Provider metadata
`provider`	string \| null	Provider name
`provider_id`	string \| null	Provider job ID
`created_at`	string \| null	ISO 8601 timestamp

Example

{
  "data": {
    "id": "550e8400-e29b-41d4-a716-446655440000",
    "url": "https://storage.example/speech.mp3",
    "status": "SUCCESS",
    "media_type": "AUDIO",
    "meta": {}
  },
  "error": null
}

If status is PENDING, poll Jobs.