Sub-styles and prompts to start from
| Sub-style | Prompt to start from | Mood setting |
|---|---|---|
| Classic study beat | Lo-fi hip hop, dusty drums, mellow rhodes piano, vinyl crackle, 75 bpm, no vocals | chill |
| Jazzhop | Jazzy lo-fi, muted upright bass, brushed drums, warm electric piano chords, late night, 80 bpm | romantic |
| Rainy day | Lo-fi with soft rain texture, slow acoustic guitar loop, tape hiss, sleepy kick and snare, 68 bpm | sad |
| Sleep and ambient | Very slow lo-fi, pad synth, sparse piano notes, no drums, soft hiss, calm and spacious | chill |
| Chillhop with vocal hook | Lo-fi chillhop, laid-back drums, warm bass, whispered female vocal hook, dreamy | romantic |
| Synth lo-fi | Lo-fi synthwave, analog pads, slow gated drums, nostalgic arcade mood, 85 bpm, instrumental | dark |
Copy a prompt into the box above and change the subject, instruments or tempo. Do not name a real artist: describe the voice and the sound instead.
Lyric and structure tips for this genre
- Lo-fi works because it is repetitive. Ask for a loop-friendly feel and few sudden changes so the track can sit under a 2-hour video.
- Name two or three instruments, not ten. Rhodes, muted guitar or soft piano plus a dusty drum is enough; extra layers make the mix busy.
- Name the texture: vinyl crackle, tape hiss, rain, cafe noise. These small words change the character more than the genre label does.
- Keep the tempo between 60 and 90 bpm and say it in the prompt. Below 70 reads as sleepy, above 85 reads as upbeat chillhop.
- Vocals are optional. If you want one, ask for a short whispered or sampled-sounding phrase and use the lyrics tab with four lines at most.
- To make a longer stream, generate two tracks with the same prompt and crossfade them in your editor.
Hear the engine first
Lo-fi Study is an unedited lo-fi track from the same engine; Late Shift is a jazzy hip hop track that shows the dusty-drum side of it.
Every generation returns two versions from MiniMax Music v2 and ACE-Step, which often differ in arrangement.
What to use it for
| Use | What you make | Setting |
|---|---|---|
| Study and focus streams | Loop under a 24/7 or long video | Instrumental, 70 bpm, no melody peaks |
| Sleep and meditation videos | Slow tracks with no drums | Ambient pad and sparse piano |
| Podcast and vlog beds | Quiet under a voice | Soft drums, one chord instrument |
| Twitch starting-soon screens | Short looping track | Jazzhop, upbeat 85 bpm |
| Phone and notes app ringtone-style clips | Short personal use | Any prompt, free tier |
Licence and pricing
| Option | Price | Licence |
|---|---|---|
| Free | $0 | 1 generation a day (2 songs), up to 2 min, personal use only |
| Small pack | $4.90 once | 200 credits, about 40 songs, commercial licence |
| Large pack | $12.90 once | 600 credits, about 120 songs, commercial licence |
| Pro | $9.90 a month or $99 a year | 450 credits every month, songs up to 3 min, commercial licence |
The licence follows the credits used to make a song. A song made on the free allowance stays personal even if you buy later, so make the final version after you buy.
Frequently asked questions
Can I make a lo-fi beat with no vocals?
Yes. Keep the vocal on none and describe the instruments. You get two instrumental versions of each prompt as MP3 downloads.
Can I use the lo-fi tracks on a monetised YouTube stream?
Yes, if you make them with purchased credits or while on a paid plan; those come with a commercial licence that stays valid. Free tracks are personal use only.
How do I make a lo-fi track loop?
Ask for a steady, repetitive groove with few changes, then crossfade two generated tracks or repeat one in your editor. We do not guarantee a perfect loop point on a single file.
What tempo is lo-fi?
Most lo-fi sits at 60 to 90 bpm. Put the number in your prompt.
Is it really free?
Free tracks are limited to a daily allowance and are for personal use. Paid packs start at $4.90 once, no subscription.
Can I get a lo-fi sound like a specific producer?
Describe the sound instead, for example dusty drums and warm rhodes. We do not support imitating a named artist's voice or catalogue.