Try Seedance 2.5 Reference-to-Video from ByteDance →
Sign InStart Building

Task History

Click to see output list

No tasks yet

Go to Models
Video GenerationActive

wiro / mmaudio

mmaudio

bywiro

Generate synchronized audio for videos using AI. Enhance video content with matching sound effects or create custom audio from text descriptions.

Video to VideoFast InferenceText To Audio
Model ID
mmaudio
Provider
wiro
Updated
1773440828
wiro playground—wiro/mmaudio
Reset to defaults

Upload a video file to generate synchronized audio. The model will analyze visual content and create matching sound effects. Maximum 30 seconds. If no video is provided, the model switches to text-to-audio mode.

Alternative to file upload. Provide a direct URL to a video file. Leave empty if uploading a file directly.

Describe the audio you want to generate. For video input: enhances the generated audio with specific sounds or leave empty. For text-to-audio mode: required description of the audio to create.

Describe sounds to avoid in the generated audio. Helps filter out unwanted elements like background noise, distortion, or specific sounds you don't want.

Sample outputs
Click To See More Samples
Updated 1773440828
# MMAudio - Video to Audio Generator ## Overview MMAudio is an AI-powered tool that generates synchronized audio for video content. It analyzes visual content and creates matching sound effects, or creates custom audio from text descriptions. The model supports both video-to-audio synchronization and text-to-audio generation modes. ## What you can build - Enhanced video content with synchronized sound effects - Custom audio for creative projects - Automated audio generation for multimedia productions - Audio enhancement for existing video materials ## Inputs - **Input Video**: Upload a video file (maximum 30 seconds) for synchronized audio generation - **Input Video URL**: Alternative method using a direct video URL link - **Prompt**: Text description of desired audio output - **Negative Prompt**: Specify unwanted sounds or audio elements ## Outputs - Generated audio file synchronized with input video - Custom audio files created from text prompts ## Recommended settings - For best results, use clear video content with distinct visual elements - Set inference steps between 30-50 for balanced quality and speed - Adjust strength parameter between 4.0-5.0 for optimal prompt adherence ## Limitations - Video processing is limited to 30 seconds maximum - Audio quality may vary depending on video complexity - Text-to-audio mode requires descriptive prompts for best results ## Safety & compliance - All audio generation occurs locally on the platform - No personal data is stored or transmitted during processing - Users are responsible for ensuring content complies with local regulations

Example prompts

Great starting points for mmaudio.

ocean waves crashing on the beach, seagulls, windVideo Generation
horse galloping, hooves on groundVideo Generation
lightning storm, thunder, strong wind, rainVideo Generation
thunderstorm with heavy rain and distant thunderVideo Generation

API quick start

Run mmaudio with a single API call.

POST https://api.wiro.ai/v1/Run/wiro/mmaudio
{
  "prompt": "ocean waves crashing on the beach, seagul…",
  "inputVideo": "https://your-cdn.com/input.mp4",
  "inputVideoUrl": "...",
  "negativePrompt": "bad, blurry"
}
View full API docs