Models/Video to Audio

Video to Audio API — Generate Sound for Silent Video

Live2 models

Muapi's video-to-audio API analyzes a silent video and generates a matching ambient track, foley, or sound effect synchronized to the on-screen action — no manual sound-design pass required. The same MMAudio family can also generate a standalone sound effect from a text prompt alone. Both workflows use the unified REST API and cost $0.001 per second of generated audio.

2 of 2 models
MMAudioVideo to Audio

MMAudio Video to Audio

Analyze a silent video and generate a matching ambient track, foley, or sound effect synchronized to its visible action from a text prompt.

Video in, audio out
$0.001 / second
Try Model
MMAudioText to Audio

MMAudio Text to Audio

Generate a standalone sound effect or ambient clip from a text prompt alone, with no video input required.

Text in, audio out
$0.001 / second
Try Model

What is a Video to Audio API?

A video-to-audio API takes a silent video clip and generates a soundtrack that matches what is happening on screen — footsteps, wind, machinery, ambience, or other foley — guided by a text prompt rather than a manual sound-design pass.

Muapi exposes MMAudio Video to Audio for that workflow, plus MMAudio Text to Audio for generating a standalone sound effect or ambient clip without video input. Both use the same submit-and-poll REST pattern used across the platform and bill by the duration of the generated audio.

Key Capabilities

Video-Synchronized Sound Generation

MMAudio Video to Audio analyzes the input video and generates audio timed to visible action, guided by your prompt describing the desired sound.

Standalone Sound Effects

MMAudio Text to Audio generates a sound effect or ambience clip from text alone, useful for building a reusable sound library.

Adjustable Output Duration

Both models accept a duration parameter, defaulting to 8 seconds, so the generated clip can match your video or editing timeline.

Silent-Clip Pipeline

Add a matching soundtrack to an AI-generated video that has no audio by sending the clip directly to MMAudio Video to Audio.

Low Cost Per Second

Both MMAudio models bill at $0.001 per second of output audio, so the default 8-second clip costs $0.008.

Pairs with Speech and Music Tools

Combine generated ambience or effects with Text to Speech for dialogue or Suno for music to build a complete video soundtrack.

Video to Audio Model Comparison

ModelInputPriceBest For
MMAudio Video to AudioVideo + text prompt$0.001 / second ($0.008 for 8s default)Adding synchronized sound to a silent video clip
MMAudio Text to AudioText prompt only$0.001 / second ($0.008 for 8s default)Standalone sound effects and ambience

How to Generate Audio for Video via API

  1. Pick a model. Use Video to Audio when you have a silent clip to score, or Text to Audio for a standalone sound effect.
  2. Write a prompt. Describe the sound you want, such as heavy rain on a tin roof with distant thunder.
  3. Submit the request. Call POST /api/v1/mmaudio-v2/video-to-video with prompt, video_url, and optional duration. Text to Audio uses /api/v1/mmaudio-v2/text-to-audio without video_url.
  4. Poll for completion. Check GET /api/v1/predictions/{request_id}/result until status is completed, then download the audio.
  5. Mux into your video. Combine the generated audio track with your original silent clip in your editing pipeline.
curl -X POST https://api.muapi.ai/api/v1/mmaudio-v2/video-to-video \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "prompt": "Heavy rain on a tin roof with distant thunder",
    "video_url": "https://example.com/silent-video.mp4",
    "duration": 8
  }'
import requests

response = requests.post(
    "https://api.muapi.ai/api/v1/mmaudio-v2/video-to-video",
    headers={"x-api-key": "YOUR_API_KEY"},
    json={
        "prompt": "Heavy rain on a tin roof with distant thunder",
        "video_url": "https://example.com/silent-video.mp4",
        "duration": 8,
    },
)
request_id = response.json()["request_id"]

result = requests.get(
    f"https://api.muapi.ai/api/v1/predictions/{request_id}/result",
    headers={"x-api-key": "YOUR_API_KEY"},
)
print(result.json())

Frequently Asked Questions

What is the Video to Audio API?

Muapi's video-to-audio API analyzes a silent video and generates a matching soundtrack through MMAudio Video to Audio, guided by a text prompt describing the desired sound.

Can I generate a sound effect without a video?

Yes. MMAudio Text to Audio generates a standalone sound effect or ambient clip from a text prompt alone, with no video input required.

How much does video-to-audio generation cost?

Both MMAudio models cost $0.001 per second of generated audio. The default 8-second clip costs $0.008.

How long can the generated audio be?

Duration is a request parameter and defaults to 8 seconds. Set it to match your video length or desired clip length within the model's supported range.

Does this generate speech or dialogue?

No. MMAudio is built for ambience, foley, and sound effects. Use Muapi's Text to Speech API for speech and dialogue.

Can I get Video to Audio API access right now?

Yes. Sign up at muapi.ai, create an API key from your dashboard, and start generating audio immediately — no waitlist required.