Prompts and the tracks they produced
Audio has no thumbnail. Each entry is the text that was typed and the waveform of what came back.
- Prompt
“Warm indie folk, acoustic guitar and brushed drums, hopeful, mid-tempo, female lead vocal”
1:20 - Prompt
“Warm lo-fi beat, 80bpm, dusty Rhodes and brushed drums, soft vinyl crackle”
2:05 - Prompt
“Bright synth pop, mid-tempo, punchy drums and a lifted chorus, female lead vocal”
1:30 - Prompt
“Cinematic orchestral build, low strings into taiko drums, tense, resolving on a single held chord”
1:00
The model decides the shape of the brief
Pick the model first
Minimax Music V2 writes songs with vocals and takes the description and the lyrics as two separate fields. Seed Audio 1.0 writes instrumental beds and textures from the description alone. Switching model is how you get an instrumental — there is no vocals toggle.
Write the brief
Genre, instrumentation, mood and tempo, in plain language, up to 2000 characters. The product's own test: "Warm lo-fi beat, 80bpm, Rhodes and brushed drums, for a product montage" is a brief; "chill music" is not. On Minimax the lyrics field is required alongside it — leave it empty and the button reads "Lyrics needed".
Choose the file, then generate
Both models offer a format. Minimax adds a sample rate and a bitrate. The run is charged flat whatever the track's length, the finished file lands in your workspace library, and failed runs are refunded.
What Music takes, and what it returns
What Music needs from you
- Description
- Text, up to 2000 characters — required on both models
- Lyrics
- Required on Minimax Music V2, as a separate field. Seed Audio takes none
- Price
- 91 credits flat on Minimax, 564 flat on Seed Audio — the same either way, whatever the track's length
Export specifications
- Formats
- MP3 or FLAC on Minimax, MP3 or WAV on Seed Audio
- Sample rates
- 16000, 22050, 24000, 32000 or 44100 Hz — Minimax only
- Bitrates
- 32000, 64000, 128000 or 256000 — Minimax only
- Price
- Flat on both models: 91 credits on Minimax, 564 on Seed Audio
- Failed runs
- Refunded
- Delivery
- Rendered audio file in your workspace library
The brief, the words, and the file you get
Two models with two different shapes of control: one takes a description and a lyric sheet and sings; the other takes a description and plays.
-
Description
Up to 2000 characters
Genre, instrumentation, mood and tempo. Required on both models, and the only handle on arrangement — "Warm lo-fi beat, 80bpm, Rhodes and brushed drums, for a product montage" is a brief; "chill music" is not.
-
Lyrics
Required on Minimax Music V2
A separate field from the description, and what the voice sings. Minimax needs both — an empty lyric sheet is a rejected request, and the generate button says so before you spend anything.
-
Format
MP3 or FLAC on Minimax, MP3 or WAV on Seed Audio
The one control both models share, with a different list on each. FLAC and WAV for anything going into an edit.
-
Sample rate
16000, 22050, 24000, 32000 or 44100 Hz
Minimax only. 44100 for anything going into an edit; the lower rates are for previews and voice-over beds.
-
Bitrate
32000, 64000, 128000 or 256000
Minimax only. Sets the size and fidelity of the exported file. 256000 is the ceiling.
Engine choice
Two ways to run it. Pick the one that matches your source.
-
Minimax Music V2
VocalsSongs with vocals. Takes a description and the lyrics separately, and needs both. 91 credits flat, whatever the track's length.
Songs with a sung line
-
Seed Audio 1.0
InstrumentalByteDance's model. Instrumental beds and textures from a plain description, with no lyric sheet to write. 564 credits flat.
Instrumental beds and textures
Built for people who need a track and don't have one
-
Video editors
Score a cut without licensing a library track that a thousand other videos already use.
-
Game and app developers
Generate menu and level beds to a brief on Seed Audio, and re-render them when the brief changes.
-
Podcasters
Make an intro and outro that belong to the show rather than to a stock catalogue.
Often used with
Built for Growth at Every Stage
Every plan includes every model and every feature. Plans only change how many credits you get and how many generations run at once.
- Credits refresh monthly
- Top-up additional credits anytime
- Unused credits don't roll over
Not ready for the commitment?
FAQ
Frequently asked questions
Everything you need to know about Music on BeHooked.
How do I get an instrumental?
Switch model. Seed Audio 1.0 writes instrumental beds and textures from a plain description and takes no lyrics at all. You cannot get an instrumental out of Minimax Music V2 by leaving its lyrics field empty — Minimax needs both the description and the lyrics, and an empty lyric sheet is a rejected request. The generate button reads "Lyrics needed" rather than letting you spend anything on it.
Does it generate vocals?
Minimax Music V2 does. It takes the description and the lyrics as two separate required fields, and what you write in the lyrics field is what gets sung. Seed Audio 1.0 is the instrumental model and has no vocal line.
How much does a track cost?
Minimax Music V2 is 91 credits flat and Seed Audio 1.0 is 564 credits flat. Both are flat prices: a run costs the same regardless of how long the finished track is. Failed runs are refunded.
How long can the brief be?
Up to 2000 characters, and it is worth using the room. The product's own test is that "Warm lo-fi beat, 80bpm, Rhodes and brushed drums, for a product montage" is a brief and "chill music" is not — instrumentation, tempo and intended use written out plainly get you closer than a genre tag.
What formats does it export?
Format is the one control both models offer, with a different list on each: MP3 or FLAC on Minimax Music V2, MP3 or WAV on Seed Audio 1.0. Minimax adds a sample rate — 16000, 22050, 24000, 32000 or 44100 Hz — and a bitrate of 32000, 64000, 128000 or 256000. Take FLAC or WAV into an edit.
