Mureka V9 AI Song Generator
Turn written lyrics into a full vocal song with Mureka V9. Optional style guidance, one or two songs per run, mp3 or wav output, 21 credits per generated song on Epochal.
Model Guide
Mureka V9 is Mureka's lyrics-to-song model. On this page you write lyrics and the model returns a full vocal song, with an optional style prompt for genre, mood, and instrumentation. The exposed controls are number of songs (one or two) and output format (mp3 or wav). There is no clip-length or duration setting; the model decides the song length from the lyrics. Each generated song costs 21 credits on Epochal.
Start
Write lyrics, then steer the style
Lead with finished lyrics
Mureka V9 is lyrics-first. Write or paste your full lyrics. Verse, chorus, and line breaks are kept as you wrote them.
Add style guidance if you want it
Use the prompt to name a genre, mood, or instrumentation. Skip it to let the model read the lyrics alone on the first pass.
Pick count and format before generating
Choose one or two songs and mp3 or wav. Each generated song costs 21 credits, so two songs cost 42 credits.
Decision
How parameters change the result
Impact: The required input. The model turns your lyrics into a sung track. It does not build a song from a style prompt alone.
Recommendation: Write complete lyrics with the structure you want. The model follows your line breaks and section order.
Impact: Optional style guidance that steers genre, mood, and instrumentation. Empty lets the model infer style from the lyrics.
Recommendation: Add it when you want a specific genre or mood. Leave it empty first to see where the lyrics lead, then tighten.
Impact: One or two songs per run. Two gives you a variation to compare against the first.
Recommendation: Use one while the lyrics and style are still changing. Use two to compare takes on a stable lyric.
Impact: mp3 or wav. wav is lossless and larger; mp3 is smaller and fine for most drafts and sharing.
Recommendation: mp3 for iteration and sharing. wav when you need a lossless file for further production.
Cost
Credits per song
| Workflow | Credits per song |
|---|---|
| Lyrics to song | 21 |
Method
How to judge the first result
Check the vocal and the lyric: The clearest signal is whether the vocal sings your lyrics clearly, with the verse and chorus structure intact. If words are dropped or blurred, tighten the lyrics before you change the style.
Then judge the style match: If you set a style prompt, check whether the genre, mood, and instrumentation match. If they do not, make the prompt more specific about tempo, instruments, or a reference genre.
- Run one song while the lyrics and style are still changing, then switch to two to compare takes.
- There is no duration control. If a song feels cut short or too long, edit the lyrics rather than look for a length setting.
- Keep the style prompt short and concrete. Long prompts dilute the direction.
FAQ
Mureka V9 frequently asked questions
- What does Mureka V9 do on this page?
- It turns your lyrics into a full vocal song. Lyrics are required, and an optional style prompt steers genre, mood, and instrumentation.
- Can I set the song length or duration?
- No. Mureka V9 has no duration or clip-length parameter. The model decides the song length from the lyrics. Edit the lyrics if you want a shorter or longer result.
- What controls are exposed in the workbench?
- Lyrics (required), an optional style prompt, number of songs (one or two), and output format (mp3 or wav).
- Do I need musical skills to use it?
- No. You write the lyrics and the model handles vocals and arrangement. Mureka positions the model as usable without music production experience.
- How much does a run cost?
- Each generated song costs 21 credits on Epochal. Generating two songs costs 42 credits.
- Can I use the output commercially?
- Mureka's published terms permit commercial use of generated output, but confirm the current terms on Mureka's site for your specific use case before you publish or monetize.