What You'll Create

In this guide, you will create a short background music candidate with VMAI AI Music Generator.

By the end, you will be able to:

choose model quality for a quick preview;
use Create mode for instrumental background music;
write a music prompt within the 500-character limit;
choose Auto or Manual duration;
tune genre, mood, BPM, key, and time signature;
check Credits and sign-in requirements before generating;
review, play, and download the result when it is ready.

Note This tutorial is based on the public VMAI AI Music Generator page and the current implementation. Actual generation depends on sign-in, plan, available Credits, selected model quality, and duration. Always check the on-screen guidance before generating.

Before You Start

What you need

The approximate length and mood of the video that needs background music.
A VMAI account. The music generation start API verifies a signed-in user.
Credits for generation. Check the estimate and remaining balance on the page.
A short first prompt. The music description field accepts up to 500 characters.

When should you use this tool?

Situation	Recommended approach
You need calm music behind a product demo	Generate instrumental music designed to stay low under narration.
You need a short mood for a short-form intro	Try a short manual duration or loop workflow.
You want to compare several moods quickly	Use Lite or Standard for multiple candidates.
You want to transform existing audio	Use Cover or Repaint only after checking reference-audio rights.

Step 1: Open AI Music Generator

Open AI Music Generator.

VMAI AI Music Generator screen showing model quality, generation mode, music description, and duration settings

AI Music Generator screen: choose model quality, generation mode, music description, and duration settings in one flow.

It helps to read the page in four sections.

Area	What it does
Model Quality	Choose Lite, Standard, Pro, or Studio with different quality and Credit multipliers.
Generation Mode	Choose Create, Cover, Repaint, or Loop.
Describe Your Music	Write the scene, genre, instruments, and mood.
Music Settings	Set duration, advanced music constraints, and Credit information.

Do not try to create the final track immediately. Treat the first pass as a 10–30 second direction check.

Step 2: Start with preview-friendly model quality

At the top, choose Model Quality.

For the first test, Lite or Standard is usually enough. Lite is the fastest, lowest-Credit preview option. Standard is useful when you want a candidate closer to the final sound. Pro and Studio target higher quality, but they also have plan requirements and higher multipliers.

Model	UI description	Use it first when
Lite	Fast · 1×	You only need to test genre and mood.
Standard	Balanced · 2×	You want a more realistic candidate without going all in.
Pro	High quality · 4×	You are refining a final candidate for prompt adherence and audio quality.
Studio	Original · 7×	You need the highest-quality candidate and can accept longer generation time.

Tip Instead of starting with Pro or Studio for a long track, find the direction with Lite or Standard first, then regenerate only the best candidate at higher quality.

Step 3: Use Create mode for background music

For new background music, choose Create mode. This mode generates music or songs from a text description.

In this tutorial, keep Add Vocals off because the goal is music behind narration. Turning vocals on opens the lyrics workflow and changes the generation into a song with AI vocals. For background music, instrumental is simpler and safer.

Understand the other modes like this.

Mode	Use case	What else you need
Create	Generate new music from a prompt	Music prompt
Cover	Make a cover from reference audio	Reference audio you have the right to use
Repaint	Transform the style of existing audio	Reference audio you have the right to use
Loop	Create loopable music	A target loop length and mood

Cover and Repaint require reference audio. If you are using team or client files, check permission before uploading.

Step 4: Write the music prompt

Enter the background music direction in Describe Your Music. The current UI limits the prompt to 500 characters.

A useful prompt includes four things.

Video context: where the track will be used.
Mood: calm, bright, tense, cinematic, playful, and so on.
Instruments or texture: piano, synth pads, drums, strings.
What to avoid: no vocals, no heavy drums, not too dramatic.

Example:

text
Calm background music for a product tutorial, soft synth pads, light percussion, not too fast, no vocals, bright and stable mood that does not compete with narration.

You can also test a Korean prompt:

text
차분한 제품 튜토리얼 뒤에 깔 배경음악. 부드러운 신스 패드와 가벼운 퍼커션, 너무 빠르지 않은 템포, 보컬 없음, 설명 음성을 방해하지 않는 밝고 안정적인 분위기.

Watch out Broad prompts such as “cool music” or “trendy music” are hard to judge. Include the video scene and what the music should avoid.

Step 5: Choose Auto or Manual duration

In Music Settings, choose duration.

Use Auto when you are still exploring. Auto lets AI decide the length based on the prompt and lyrics, then adjusts Credits after generation.

Use Manual when the video length is already known. For a 30-second short, generate near 30 seconds. For a 60-second product demo, start near 60 seconds. Manual duration caps depend on model quality.

Situation	Duration approach
Direction check	Auto or a short manual duration
Shorts and reels	Generate slightly longer than the edit, then trim
Tutorial	Generate several short scene-based candidates
Repeating background	Consider Loop mode and the repeat point

Longer tracks take more time and make candidate comparison more expensive. Start short, then make the final candidate longer.

Step 6: Narrow the result with advanced settings

Open Advanced Settings to tune genre, mood, BPM, key, and time signature.

You do not need to fill every field. If the result is too broad, lock one or two settings at a time.

Setting	When to use it
Genre	You know the direction, such as lo-fi, cinematic, or electronic.
Mood	You need a clearer emotional target, such as calm, uplifting, or epic.
BPM	The track needs to match the pace of cuts or motion.
Key	You need consistency with a brand sound or another track.
Time signature	The rhythm matters, such as 3/4 or 6/8.

If vocals are enabled, also review vocal style and vocal language. For this background-music workflow, keep vocals off and start by tuning genre, mood, and BPM only.

Step 7: Check Credits, then generate

Review Credit Info before generating.

If you are signed out, the page shows that login is required. When signed in, the page shows available Credits, rate per 30 seconds, and estimated total. The estimate changes with model quality, mode, and duration.

Before clicking generate, check this list.

Add Vocals is off.
The prompt is under 500 characters.
Auto or Manual duration matches the video goal.
Model quality and multiplier are appropriate.
If using Cover or Repaint, reference-audio rights are clear.
Estimated Credits and available Credits have been checked.

When ready, click Generate Music. The page shows progress while the track is being created, then lets you play the result.

Step 8: Review the result like an editor

When the result is ready, do not ask only “Is this good music?” Ask “Does this fit the edit?”

Check	How to test
Narration conflict	Play it under the voiceover.
Length	Check the intro, ending, and cut points.
Repetition	Listen for loops becoming obvious too soon.
Mood	Make sure the music does not overpower the message.
Reuse	Save notes if it could become a repeatable brand sound.

Download the track you like, and save the prompt and settings with it. If you keep only the audio file, it will be harder to recreate the same mood later.

Next steps

After creating a background music candidate, you can continue with these workflows.

To create the full video, start from New Video with your idea or script.
If you also need narration, read Create a Voiceover with AI TTS Voice Generator.
To decide what to check before generation, read Before You Generate AI Music: 6 Decisions That Matter More Than the Prompt.

Create Background Music with AI Music Generator