How to Make Royalty-Free Background Music with Loudly
Loudly generates original music from a text description: genre, mood, tempo, instruments. You get a downloadable track in roughly 15 to 30 seconds, and the paid plans include a licence that covers commercial use including monetised YouTube videos.
That licence is the reason to care. Free AI music generators often produce tracks you cannot legally monetise, which is a problem when the whole point is a channel with ads on it.
How do you write a prompt that produces usable music?
Describe genre, mood, tempo, and one or two instruments. Skip adjectives like “amazing” and “professional” — they change nothing and push the model toward generic output.
A prompt that works:
Lo-fi hip hop, 85 BPM, mellow, Rhodes piano, soft vinyl crackle,
no drums in the first 15 seconds.
A prompt that wastes a generation: “good upbeat background music for my video”. No tempo, no genre, no instrumentation. You get a template.
Tempo matters most. 70 to 90 BPM reads as calm, 100 to 120 as neutral, above 130 as energetic. If you are unsure, set the tempo and let the mood follow.
How long should the track be?
Match the length to your edit, not the other way round. Generating a 3-minute track for a 45-second clip wastes credits.
Two approaches:
- Generate short, loop twice. A 60-second track looped with a crossfade covers most product videos and sounds intentional.
- Generate long for talking-head content. A 3-minute track avoids a noticeable loop point under dialogue.
Loudly has a stem export on paid plans, which matters more than track length. Pulling drums or vocals out of the mix is what makes a track sit under a voiceover without fighting it.
How do you use the remix feature?
The catalogue of royalty-free tracks can be remixed: change tempo, swap instruments, alter the structure. This is faster than generating from scratch and gives a more coherent result, because you start from a professionally produced track.
When remixing, change one thing at a time. Tempo plus instrumentation plus key becomes unpredictable. Tempo alone is reliable.
What does it cost and what are the limits?
Loudly runs a free tier with limited downloads and a monthly subscription for commercial use and stem exports. Check the current pricing page; plan tiers shifted through 2026.
Limits worth knowing before you commit:
- Free tier downloads may carry attribution requirements, and some tiers restrict monetised use.
- Vocal generation is weaker than instrumental generation. For anything with lyrics, expect to regenerate three or four times.
- Generated tracks are short by default; extending requires generating again rather than trimming.
Where does AI music still fall short?
Structure. Generated tracks often have a flat energy curve — no build, no drop, no resolution. For a 30-second ad that is fine. For a 4-minute narrative video, you will be editing music rather than using it.
Also transitions. Cutting between two AI tracks rarely works without a manual crossfade, and abrupt key changes stand out.
Who is this for?
Faceless YouTube channels, podcast intros, game prototypes, and anyone producing more than five videos a month who cannot pay a composer per track.
Not for releases. If you are distributing music to Spotify or selling it, use a human composer or a legally airtight library. The licensing terms of AI generation are still being tested in court.
Quick Summary
- Loudly generates original tracks from a prompt in roughly 15 to 30 seconds; paid plans license commercial and monetised use.
- Prompt with genre, BPM, mood and two instruments; tempo is the setting that most changes the result.
- Generate 60 seconds and loop twice for short edits; use 3-minute tracks under dialogue.
- Remix one parameter at a time — tempo plus instrumentation plus key is unpredictable.
- Instrumentals are stronger than vocals, and generated tracks often lack a build-and-release structure.
