Write Lyrics, Generate the Song, Then Build Its Music Video
Music Studio takes a song from words to a finished video in one place. Start with lyrics: have the AI writer draft them in a form you choose, such as verse and chorus, AABA, 12-bar blues, rap, ballad, duet, anthem or through-composed, with the rhyme scheme you want. Or paste lyrics you already wrote. Or skip the writing and upload a finished track from Suno, Udio or your own DAW.
Then generate the song. The studio runs Mureka, Google Lyria 3 Pro, MiniMax Music 2.6 and Eleven Music v2 side by side and shows you every provider request before a token is spent. Each track comes back with word-timed lyrics you can adjust on a millisecond ruler and export as LRC or SRT. You can also pull stems, extend the track, make a seamless loop and generate cover art.
The last step is the music video. A director model plans shots across the song's sections. Each shot gets image and video takes you can reroll or swap, and the render lays your clips over the song as one audio bed. Everything is available through the API with per-call token prices, so developers and AI agents can run the same pipeline.
Open Music Studio on Reality Fabricator
What you can do
- AI lyric writer with song forms from verse-chorus to 12-bar blues and a chosen rhyme scheme
- Generate on Mureka, Google Lyria 3 Pro, MiniMax Music 2.6 and Eleven Music v2 side by side
- Word-timed lyrics on a millisecond ruler, exported as LRC or SRT
- Stems, track extensions, seamless loops and AI cover art for every song
- A director model plans the music video shot by shot with rerollable takes
- Every provider request is shown before you spend a token
How it works
- Draft lyrics with the AI writer, paste your own, or upload a finished track.
- Pick the music models to run and review each provider request before generating.
- Edit the word timings, export LRC or SRT, and add stems, extensions or cover art.
- Let the director plan shots, reroll any take, and render the music video over the song.
Frequently asked questions
Which music models does Music Studio use?
Mureka V9, O2 and V9.5, Google Lyria 3 Pro, MiniMax Music 2.6 and Eleven Music v2. You can run them side by side on the same lyrics and compare the results. The exact request sent to each provider is shown before any token is spent.
Can I use a song I made somewhere else?
Yes. Upload a finished track from Suno, Udio or your DAW. It gets the same word-timed lyrics, LRC and SRT export, stems, extensions, loops and cover art, and you can build a music video for it.
How do timed lyrics work?
Every generated or uploaded track is aligned so each word has a timestamp. You can nudge timings on a millisecond ruler and export the result as an LRC or SRT file for lyric displays or captions.
How is the music video made?
A director model reads the song's sections and plans a shot list. Each shot gets image and video takes that you can reroll or swap. When you are happy, the render lays the clips over the song as a single audio bed and outputs the finished video.
Is there an API?
Yes. Every stage, from lyrics to video render, is available through the same API with per-call token prices, so developers and AI agents can drive the pipeline without the browser.