Which MUSID workflow should I use for my music video?
Choose between Video Agent, individual video clips, Talking Avatar, and Studio. Learn what to upload, what to review, and how to plan credits.
MUSID helps musicians and creators make AI music videos from songs. Use Video Agent for a song with multiple scenes, AI Video Generator for individual clips, Talking Avatar for a portrait driven by audio, and Studio to arrange existing clips. You can also generate music and scene images separately.
Choose by the result you need
| What you want to make | Start here | What to prepare | What to check |
|---|---|---|---|
| A music video for your own song or an exported Suno track | Video Agent | Audio you have permission to use, a visual idea, and optional character references | Storyboard, reference images, each scene take, and the final merge |
| One scene, a visual experiment, or a promotional clip | AI Video Generator | A prompt and any references supported by the selected model | Motion, composition, duration, and aspect ratio |
| A portrait speaking or performing to supplied audio | Talking Avatar | A portrait image, audio, and a performance prompt | Mouth movement, expression, face visibility, and audio timing |
| A sequence assembled from clips you already have | Studio | Your video clips and an audio track | Clip order, trim points, framing, and the exported result |
| A song before starting the video | AI Music Generator | A style description or your own lyrics | The track, lyrics, and suitability for your project |
| A scene image, character reference, or cover | AI Image Generator | A visual description and supported references | Composition, character details, and visible text |
How do I turn my song into a multi-scene music video?
- Prepare the audio. Upload your recording or an exported song file to Video Agent. Select a short section for an initial test, or choose the full track when you are ready for a larger project.
- Describe a visual direction. Include a subject, setting, mood, and camera style. For example: “A lone singer in an empty railway station at night, blue lighting, slow camera movement, alternating portrait and wide shots.”
- Review the storyboard. Check scene prompts and generate reference images. Adjust the composition and character appearance before generating video takes.
- Generate and watch the clips. Check each scene with your audio. Revise a prompt or generate another take when motion, faces, or timing need work.
- Merge and review the whole video. Watch the full export before publishing. Use Studio if you want to rearrange or trim clips on a timeline.
For the controls and scene workflow, see the Video Agent guide. For writing individual shot prompts, see the AI Video Generator guide.
Which workflow should I use for lip sync?
Start with Talking Avatar when you have a portrait and want it to perform to supplied audio. Other video models have different audio and reference options; check the controls for your selected model.
A multi-scene music video does not automatically include accurate singing in every shot. Face angle, movement, audio clarity, and the model can affect mouth movement. Test a short passage with a clearly visible face, then inspect the result with sound before committing to a longer performance. Wide shots and scenery can be useful where a close-up performance is unnecessary.
How should I budget credits for a complete song?
A complete video may require scene images, several video clips, and additional takes. Its total cost depends on the selected model, generated duration, resolution, and how many assets you generate. A plan's credit allowance or the cost of one clip is not a fixed quote for a complete music video.
Check the current plans and credits and the charges displayed for your selected settings. Test one representative scene first, record what you spend on its images and video takes, and use that result to plan the remaining scenes. Leave room for revisions.
Generation time also varies with model, duration, scene count, and provider queues. Plan time for review and merging as well as generation.
Can I publish a video made from a Suno song or another recording?
You can upload an exported audio file to create visuals, but you should first confirm that you have permission for your intended use of the recording and lyrics. Check the current MUSID plan, service terms, and relevant model or source-service terms. Also check permission for any supplied images or likenesses.
A paid plan does not clear third-party rights or guarantee acceptance or monetization on a publishing platform. If your use is unclear, contact support@musid.ai before using the output for a client or commercial release.
What should I check before sharing the result?
- Watch the complete export with audio, not only individual scene previews.
- Check continuity between shots, recurring characters, hands, faces, and visible text.
- Check lip sync in performance shots and the timing of scene changes.
- Confirm aspect ratio and framing for the destination platform; available settings depend on the model.
- Replace or revise clips that do not meet your needs, and confirm permissions for the assets used.