Sonilo Sound Effects 1.0 | EmpirioLabs AI Docs

Sonilo · Audio Generation

POST /v1/audio/generations

Frame-accurate sound effects placed on the action in a video, or generated from a text description, with per-segment direction.

At a glance

Field	Value
Model id	`sonilo-sound-effects-1-0`
Model release date	2026-07-25
Input modalities	Text, Video
Output modalities	Audio
Context window	-
Weight precision	-
Features	sound_effects, text_to_audio, video_to_audio, audio_generation, music_generation, commercial_ready
Native inference	No
New	Yes
Supported endpoints	`POST /v1/audio/generations`
Alternate model ids	`sonilo-sound-effects-1.0`, `sonilo-sfx`, `sonilo/sonilo-sound-effects-1-0`

Pricing

Charge	Spec	Rate
Text to sound effect	per generated second	$0.0036
Video to sound effect	per generated second	$0.018
Music and effects	per generated second	$0.0288

Example request

curl https://api.empiriolabs.ai/v1/audio/generations \
  -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"model": "sonilo-sound-effects-1-0", "prompt": "warm jazz piano", "duration": 8}'

Parameters

Parameter	Type	Required	Default	Description
`prompt`	string	no	-	Description of the sound to generate. Required when generating from text alone. Optional when working from a video, where the footage leads.
`mode`	enum	no	`"auto"`	auto places effects on an attached video and falls back to text to sound effect when there is no video. sound scores the video with music AND effects as one finished track. · Allowed: `auto`, `text`, `video`, `sound`
`duration`	number	no	`8`	Effect length in seconds. Text mode only. Video and sound modes follow the timing of the source video. · Range: 0.5 – 180
`music_prompt`	string	no	-	Direction for the music bed that sits under the effects.
`sfx_prompt`	string	no	-	Direction for the sound effects placed on the action.
`output_format`	enum	no	`"wav"`	Format of the combined track. wav is 16-bit PCM, m4a is AAC, mp3 is 320 kbps. · Allowed: `wav`, `m4a`, `mp3`
`preserve_speech`	boolean	no	false	Keep the speech from the source video and carry it into the finished mix.
`ducking`	boolean	no	false	Dip the generated music under the speech so dialogue stays intelligible.
`segments`	string	no	-	JSON array of timed segment prompts for directed beats. Each entry takes start, end and prompt in seconds.
`video_url`	string	no	-	Source video to sync effects to, as a publicly reachable URL. Up to 300MB and 8 minutes. Uploading a video in the playground sets this for you.

Notes

Modes

auto places effects on the attached video, and falls back to text to sound effect when there is no video
video syncs effects to the timing of the footage
text generates an effect from a description alone
sound scores the video with music AND effects as one finished track

Controls

Supports prompt, mode, source video, 0.5 to 180 second duration in text mode, separate music and effects prompts in sound mode, wav, m4a or mp3 output in sound mode, speech preservation, ducking, and JSON segment prompts.

Limits

Source video up to 300MB and 8 minutes
Duration 0.5 to 180 seconds in text mode. Video and sound modes follow the source
Music and effects prompts, output format, speech preservation and ducking apply to sound mode only

Billing

Charged per generated second at the catalog rate for the mode used, with a 3 second minimum for text and video modes and a 10 second minimum for sound mode
A request rejected before generation is not billed.

Machine-readable schema: GET https://api.empiriolabs.ai/v1/models/sonilo-sound-effects-1-0.