🛒 AI music mass production and monetization solutions for music creators and content teams, covering AI composition, lyrics generation, BGM customization, short video soundtrack distribution and copyright revenue operations.

AI music batch creation and divine comedy matrix operation plan

Solution overview

Background and Trends

In 2025-2026, AI music generation models represented by Suno and Udio have achieved a qualitative leap. Suno V5.5 can already generate a complete broadcast-quality song including vocals, arrangement, and mixing within 10 seconds after inputting a lyric or genre description; Udio continues to iterate on its continuation and remix capabilities. At the same time, the demand for original soundtracks on short video platforms (Douyin/TikTok, Kuaishou, YouTube Shorts, Instagram Reels) has increased exponentially—a matching BGM is needed behind every popular video, and the labor costs and time cycles of traditional music production are completely unable to meet this content throughput.

The "Divine Comedy Matrix" is an operating model that was born under this gap between supply and demand: it utilizes the low-cost, high-output characteristics of the AI ​​music generation model to mass-produce brainwashing divine songs, holiday blessing songs, and hot topic soundtracks. It is put on the short video platform in the form of matrix accounts, and earns streaming media copyright shares (Music Royalties) and platform creator incentives through BGM usage. From a horizontal perspective, this model can also be extended to scenarios such as audiobook/podcast BGM customization, independent game soundtracks, e-commerce short video advertising music, etc., forming a three-in-one business closed loop of "tool end + traffic end + business end".

Target user portrait

Input-output expectations

• Single person production capacity improvement: traditional composition 1 song/day → AI assistance 10-30 songs/day (including screening optimization)
• Single production cost: Traditional production 500-5000 yuan/song → AI solution 0.5-10 yuan/song (including computing power subscription apportionment)
• Copyright revenue potential: The streaming media revenue of a single popular BGM on the short video platform can reach thousands to tens of thousands of yuan/month
• Payback period: 1-3 months under the SaaS subscription model, and the first order under the project model will be paid back immediately

Preconditions

• Have experience in using daily office software and be familiar with business processes and Internet access conditions in the field
• Understand the content ecology and traffic distribution mechanism of short video platforms
• Register relevant AI tool accounts and complete real-name authentication
• Have basic audio aesthetic and screening abilities (no professional music theory required)
• Have a certain operating budget for tool subscription (average monthly 200-1,000 yuan)

Toolchain list

Preparation

Before starting implementation, please confirm the following preparations one by one:

Account registration and authentication

• [ ] Register Suno account (suno.com), it is recommended to upgrade to the Pro version ($10/month, 500 points/day)
• [ ] Register Udio account (udio.com), Pro version ($10/month, 1200 points per month)
• [ ] Register ChatGPT account (chatgpt.com), Plus version ($20/month)
• [ ] Register ElevenLabs account (elevenlabs.io), Pro version ($22/month)
• [ ] Register/Login CapCut Account (capcut.com, free to use)
• [ ] Bind short video platform account (Douyin creation service platform, TikTok Creator Marketplace, YouTube Content ID)

Data and material preparation

• [ ] Collect recent popular BGM lists on short video platforms (TikTok Hot Songs List, TikTok Viral Songs)
• [ ] Organize holiday/hotspot calendars (Spring Festival, Valentine's Day, Christmas, college entrance examination season, graduation season, etc.)
• [ ] Prepare 10-20 sets of commonly used lyric templates and genre keyword libraries
• [ ] Establish a "List of Divine Comedy Elements": brainwashing chorus structure, repetitive melody progression, and memory point design methods

Environment configuration

• [ ] Ensure a stable network environment (it is recommended to use Western US/Asia nodes to reduce API latency)
• [ ] Prepare at least 2 short video platform accounts for matrix distribution testing
• [ ] Install CapCut desktop version (video synthesis performance is better)
• [ ] Prepare cloud storage space (Baidu Cloud Disk/Google Drive) for material archiving

Team division of labor (optional)

• [ ] Determine content planning position (hot topic selection + lyrics direction)
• [ ] Determine the AI operation position (music generation + screening + post-production)
• [ ] Determine the distribution operation position (short video publishing + data analysis + copyright settlement)

Step-by-step guide

Step 1: Hot topic selection and lyrics planning

⏱ Estimated time spent: 2-3 hours per week (can batch process selected topics for the week) 🎯 Goal: Determine the musical theme, genre direction and lyric framework to be generated this week ⚠️ Prerequisites: Have registered a ChatGPT account and organized a hotspot calendar

Operation instructions

The core of the Divine Comedy lies in "brainwashing melody + emotional resonance + hot spot association". This step solves the problem of "what song to make" - not random creation, but based on a data-driven topic selection strategy.

Specific operation path

1.1 Hotspot mining

• Open Douyin Hot Song List, TikTok Viral 50, Spotify Viral 50, and record the style keywords of the Top 20 songs this week
• Use ChatGPT (or use the Internet search function) to analyze trends: "Analyze the common musical characteristics (BPM range, mode, orchestration, lyric theme) of the Top 10 Douyin songs this week"
• Extract 3-5 reusable "hit formulas": such as "lyrical intro + electronic chorus", "Chinese style melody + modern arrangement"

1.2 Lyrics generation

• Enter prompt word template in ChatGPT:

You are a professional lyricist. Please compose a lyric based on the following requirements:
Theme: [Going home during the Spring Festival/Graduation farewell/Refreshing summer/Healing from broken love]
Style: [Pop/Pop Rap/National Style/Electronic Dance Music]
Structure: Intro 8 bars -> Verse A 16 bars -> Pre-chorus 8 bars -> Chorus 16 bars -> Interlude 8 bars -> Verse B 16 bars -> Chorus 16 bars -> Ending 8 bars
Memory points: The chorus needs to contain a repetitive memory phrase of 3-5 words.
Language: Chinese
Output format: paragraph markup structure

• Generate 3 sets of lyrics in each round and select the 1 set with the greatest potential for spreading
• Save lyrics to the topic library and mark applicable festival/scene tags

1.3 Genre parameter setting

• Match 2-3 genre parameter combinations for each lyric, for example:
  • Combination A: Pop, female vocal, 120BPM, C major, piano intro, electronic drop
  • Combination B: Chinese style, male vocal, 80BPM, A minor, guzheng, orchestral
  • Group C: Lo-fi hip hop, no vocals, 90BPM, D minor, vinyl crackle, jazz sample

Verification method

• [ ] Complete at least 10 sets of lyrics + 20 sets of genre parameter combinations every week
• [ ] ChatGPT output content has no obvious factual errors or style deviations
• [ ] Topics covering at least 3 different hotspot directions

FAQ

Q: What should I do if the generated lyrics are too "AI-flavored"? A: Add constraints - require "avoid clichés, add specific imagery (time/place/item)", and manually modify rhymes and emotional words after generation.

Step 2: AI music batch generation

⏱ Estimated time : 1-2 hours for the first batch of 20 (including waiting and screening) 🎯 Goal: Use Suno/Udio to batch produce candidate music and filter out tracks that meet release standards ⚠️ Prerequisites: Suno Pro/Udio Pro account has been recharged, lyrics and parameter configuration completed

Operation instructions

This is the core link of the entire program. Utilize the high throughput capabilities of AI music tools to replace the traditional workflow of "single-by-song crafting" with a "mass production + quality gate control" model. The key is scale, not single perfection.

Specific operation path

2.1 Suno batch generation (recommended main tool)

• Log in to Suno (suno.com) and enter the Create page
• Fill in lyrics + genre description (Style Prompt), format example:

  [Style: Pop, emotional female vocal, 128BPM, piano and strings, anthemic chorus]
  [Lyrics]
  (Intro)
  ...

• Key parameter configuration tips:
  • Turn on "Custom Mode" to customize lyrics (instead of purely automatic generation)
  • Style Prompt should be limited to 3-5 keywords. Too many keywords will dilute the style clarity.
  • Batch generation of 4-6 songs at a time (using Suno's batch generation queue)
  • Use v5.5 model for best vocal quality
• Quick audition markup after generation:
  • ✅ Passed: Clear vocals, smooth melody, and complete structure → Enter the post-production pool
  • ⏸ To be modified: overall acceptable but with obvious flaws → Regenerate or extend
  • ❌ Elimination: pitch problems, blurred vocals, disordered rhythm → delete directly

2.2 Udio supplementary generation (continuation/remix scene)

• For scenes that require "extending from existing clips", use Udio's Extend function
• Usage scenarios include:
  • Expand a 30 second demo into a full 2 minute song
  • Remix Suno output to change the orchestration style
  • Generate purely instrumental version (no lyrics) for use by BGM library
• Operation path: Upload Audio → Extend/Crop → Remix → Download

2.3 Batch output quality control access control Each candidate music must pass the following 4 checks. If it fails, it will be eliminated or regenerated:

1. Sound quality access control: no broken sound, noise floor below -50dB, sampling rate ≥ 44.1kHz
2. Structural Access: There is a clear intro-verse-chorus-outro structure
3. Vocal Access Control: The lead singer’s voice is identifiable, no synthetic mouth sounds, and the pitch deviation is < 3 cents.
4. Emotional access control: The melody emotion is consistent with the design theme (sad/joyful/excited, etc.)

Verification method

• [ ] Single batch output ≥ 20 candidate tracks
• [ ] Pass rate ≥ 30% (i.e. ≥ 6 out of 20 songs meet the release standard)
• [ ] Daily output is stable at 5-10 qualified pieces of music

FAQ

Q: What should I do if the vocals generated by Suno always have a "metallic taste"? A: Try adding descriptors such as warm analog, natural reverb, vintage in the Style Prompt; or use ElevenLabs’ vocal restoration function to process it later.

Q: How do Udio and Suno choose their main tools? A: Vocal songs with lyrics will be given priority in Suno (the vocal quality is better); purely instrumental music BGM, Remix, and fragment continuation will be given priority in Udio (the editing flexibility is higher). The two complement each other for optimal efficiency.

Step 3: Audio post-production and mixing optimization

⏱Estimated time: 5-15 minutes per song 🎯 Goal: Upgrade the original audio generated by AI to platform release standards and remove defects ⚠️ Prerequisites: Suno/Udio export completed, ElevenLabs account ready

Operation instructions

Audio directly generated by AI usually has some common problems: the human voice is not full enough, the low frequency is muddy, the high frequency is harsh, and the overall loudness is not up to standard. This step solves these problems with lightweight post-processing that requires no professional DAW experience.

Specific operation path

3.1 Vocal Restoration (ElevenLabs)

• Log in to ElevenLabs and use Voice Design or Voice Library functions
• Separate the vocal part of the AI-generated track (using Suno's own Stem Splitter or a third-party separation tool)
• Process vocals with ElevenLabs’ sound quality enhancement channel
• Key parameters: Clarity +30%, Stability 40-60%, Style Exaggeration 20%

3.2 Master-level loudness normalization

• Use the audio adjustment features of CapCut Desktop:
  • Normalize overall loudness to -14 LUFS (recommended standard for short video platforms)
  • Compressor: Ratio 3:1, Threshold -18dB, Attack 10ms, Release 50ms
  • Equalizer: Cut off infrasound below 60Hz and slightly boost 3-5kHz to increase presence.
• Export format: MP3 320kbps or WAV 16bit/44.1kHz

Verification method

• [ ] Processed audio loudness in the range -14 ± 1 LUFS
• [ ] There is no harshness or distortion when listening on the mobile phone and through headphones.
• [ ] Standard naming convention for all version files, clear archiving

FAQ

Q: How to do post-production without professional audio software? A: CapCut’s built-in audio adjustment function covers 90% of your needs. For more granular control, use the free tool Audacity.

Step 4: Short video soundtrack synthesis and distribution

⏱Estimated time: 10-30 minutes per video 🎯 Goal: Combine AI music and video materials into publishable short videos to complete matrix account distribution ⚠️ Prerequisites: Audio post-production completed, short video material ready, matrix account ready

Operation instructions

Music itself is the "ammunition", and short videos are the "battlefield". The core of this step is to embed AI music into short video content and use the communication attributes of BGM to obtain traffic and copyright revenue. Key strategies: One piece of music matches multiple videos, multiple accounts are distributed, and multiple platforms are covered.

Specific operation path

4.1 Video material preparation

• Use CapCut template library or AI to generate video footage
• Source of material:
  • CapCut template library (free for commercial use)
  • Own real-life footage
  • AI-generated graphics/animations (applicable to pure music display videos)
• Music rhythm alignment: Align the video screen switching points with the beat and chorus starting point of the music on the CapCut timeline

4.2 Video synthesis parameters

• Frame: 9:16 vertical screen (Douyin/Kuaishou/TikTok default), 16:9 horizontal screen (YouTube)
• Duration: 15-60 seconds (best for short videos), 2-3 minutes (full version on YouTube)
• Subtitles: CapCut automatic subtitles + lyrics displayed line by line (increases completion rate)
• Cover: unified visual style (text title + music name + creator logo)

4.4 Copyright registration (key actions)

• Under the Suno Pro package, all builds are yours with full usage rights
• Udio commercial authorization requires confirmation of package permissions (Pro version supports commercial use)
• Perform musician certification on the short video platform and upload the audio to the "Original Music" library
• Mainland China recommends simultaneous digital copyright registration with the "China Music Copyright Association"
• Use the Content ID system (YouTube) to monitor BGM usage

Verification method

• [ ] Publish ≥ 5 short videos (including BGM) in a single day
• [ ] Each video uses copyrighted original music
• [ ] Matrix account establishes a unified visual identification system (cover style, watermark position, description format)

FAQ

Q: What should I do if the BGM is stolen by other accounts after it is released? A: After completing the copyright registration, you can go through the platform reporting process. YouTube Content ID can automatically identify and account for accounts.

Step 5: Data analysis and copyright revenue operation

⏱ Estimated time: 2-3 hours per week 🎯 Goal: Track music usage data, copyright income, and iteratively optimize production strategies ⚠️ Prerequisites: The content has been published ≥ 7 days, and the data has been accumulated sufficiently

Operation instructions

The Divine Comedy Matrix is not a linear process of "production-release", but a closed loop of "production-release-monitoring-iteration". Only by continuously optimizing the topic selection direction and genre parameters through data feedback can we develop the ability to stably produce hot hits.

Specific operation path

5.1 Data indicator board

• Track the following core metrics:
  • BGM Usage: The number of times your music has been used as video BGM by other creators
  • Completion rate: The average completion rate of videos using your BGM
  • Copyright Income: Streaming Media Account + Platform Incentive Fund
  • Blowout rate: Proportion of tracks with usage > 1000 times

5.2 Iterative decision-making

• Adjust the next round of production direction based on data feedback:
  • High completion rate but low usage → The genre is too narrow, adjust the style to a more universal direction
  • High usage but low completion rate → The chorus memory points are not enough to brainwash, optimize the melody Hook
  • The usage of songs related to specific festivals has surged → Plan the topics for the next festival 2-4 weeks in advance

5.3 Business Realization Expansion

• Income source matrix:	Income type	Proportion estimate	Operation method
  Short video platform copyright split	40-50%	Musician certification + original music storage
  SaaS subscription (BGM customized platform)	20-30%	Build an AI BGM custom platform
  Project-made soundtrack	15-20%	Game/audiobook/podcast customization
  Commercial authorization	5-10%	Enterprise/e-commerce batch authorization

Verification method

• [ ] Issue a weekly data report every week to compare changes in various indicators last week
• [ ] Determine the topic selection direction for next week based on data (at least 3 new directions)
• [ ] Total monthly copyright revenue covers tool subscription costs (break-even confirmation)

FAQ

Q: What should I do if I post 100 videos but none of them are hits? A: Check whether the topic selection is too broad (don’t do “post everything”). Shrink to 1-2 vertical styles (such as "Chinese style healing" or "electronic rhythm"), focus on testing a small area and then expand.

Expected results

Quantitative indicators

Output list

• [ ] Qualified AI music library ≥ 200 songs (including full version + short video version + accompaniment version)
• [ ] Finished short videos with soundtrack ≥ 500
• [ ] Original music with copyright registration ≥ 100 pieces
• [ ] Operation data dashboard (including delivery effects and revenue summary)
• [ ] A copy of SOP operation manual (including prompt word template, parameter configuration table, troubleshooting guide)

Acceptance criteria

• [ ] Can stably produce 5-10 qualified AI music pieces every day
• [ ] Total fans of short video matrix account ≥ 10,000 or total monthly views ≥ 500,000
• [ ] Monthly copyright income can cover tool subscription costs
• [ ] At least 1 track with usage > 10,000 times (small hit)
• [ ] The team has the ability to independently complete the entire process from topic selection to copyright settlement.

Frequently Asked Questions and Troubleshooting (FAQ)

Q1: ​​Are there any copyright issues with AI-generated music? Will it be judged as infringement by the platform? A: Both the Suno Pro/Premier and Udio Pro packages grant the user full commercial use rights to the generated content. But before publishing, be sure to confirm whether your subscription plan covers commercial use. It is recommended to complete copyright registration and authenticate original music on the platform to avoid being misjudged as infringing content. Mainland China recommends simultaneous digital copyright filing.

Q2: Why do the vocals in songs generated by my Suno sound "plastic"? A: Common causes and countermeasures: (1) Style Prompt is too short - contains at least one style word + one vocal descriptor, such as Pop, warm female vocal; (2) Model version - make sure to use v5.5 instead of the old version; (3) Post-processing - use the audio repair channel of ElevenLabs or CapCut for loudness normalization and equalization adjustment.

Q3: How many posts can a music account post per day without being restricted? A: It is recommended that a new account should post 2-3 posts per day for the first 2 weeks, and gradually increase to 5-8 posts. Each video content should be differentiated (different screen materials, titles, hashtags), and should not be published purely to swipe the screen. The same music can be used to produce multiple videos with different images and released in a staggered manner.

Q4: How big a team is needed for Divine Comedy Matrix? Can an individual do it? A: Individuals can definitely start. Minimum configuration: 1 person (yourself) = planning + operations + distribution. It is recommended to use Notion or Feishu to create a topic library and schedule. When the stable monthly income is >$1,000, it can be expanded to a team of 2-3 people (1 person for planning, 1 person for AI operation, and 1 person for distribution operation).

Q5: How to avoid "collision" with other AI music creators (exactly the same style)? A: (1) Lyrics dimension: Manually modify the lyrics output by ChatGPT and add unique personal expression; (2) Style dimension: Mix uncommon style combinations, such as R&B x guzheng, Lo-fi x Spanish guitar; (3) Post-production dimension: Add custom sound effect layers (ambient sound, white noise, vocal samples) to create recognition.

Q6: How to activate the copyright split on the short video platform? A: Douyin: Enter the "Musician Center" → Musician Certification → Upload original audio → Turn on "Use Authorization"; TikTok: Enter the SoundOn platform (soundon.com) to register as a musician → Upload for distribution; YouTube: YouTube Studio → Content ID Statement → Associate Copyright Agent. The review cycle of each platform ranges from 3-14 days.

Q7: How do I track the revenue after my music is used a lot? A: Each platform has a backend data analysis panel for creators. It is recommended to use an aggregation tool such as DistroKid, TuneCore or RouteNote for multi-platform revenue aggregation. For domestic platforms, Douyin musicians can directly view the "Copyright Income" module in the background.

Q8: Do you need professional music production knowledge? A: No need. The premise of the design of this program is that "no knowledge of music theory is required". The core capability requirements are: aesthetic screening (being able to tell which version sounds better), batch operation efficiency, and data sensitivity. Basic knowledge of arranging or mixing is a plus, but not a requirement.

Advancement and Expansion

1. Build an AI BGM mass customization platform (SaaS)

When the music library accumulates to 1,000+ songs, an AI BGM selection platform can be built for independent game developers, podcasters, and e-commerce teams:

• Technology stack: web front-end (React/Vue) + music classification retrieval (style/mood/BPM/duration) -Business model: Monthly subscription ($9.99-49.99/month), unlimited downloads
• Differentiation: All tracks adopt "AI generation + manual screening" dual quality control, and come with a copyright use certificate

2. Multi-language music market expansion

Utilize the multi-language vocal synthesis capabilities of ElevenLabs to translate and regenerate Chinese lyrics into English, Japanese, Spanish, and Arabic versions, entering the short video ecology of the corresponding language market:

• English market: TikTok + YouTube Shorts + Instagram Reels
• Japanese and Korean markets: LINE Music + YouTube Japan
• Southeast Asia market: TikTok Southeast Asia

3. Live broadcast of real-time AI soundtrack

Use Mubert's API or Suno's real-time generation interface to build a real-time AI soundtrack system for the live broadcast room:

• Automatically switch BGM according to the atmosphere of the live broadcast room (bringing goods/chatting/games/singing)
• Audience interaction driven: rewards trigger specific melodies or sound effects
• Suitable for game live broadcast, voice chat room, virtual anchor and other scenarios

4. Integrate with audiobook/podcast ecosystem

Archive the AI-generated soundtracks according to emotional tags (healing/suspense/epic/warmth) and provide them to creators on audiobook production platforms or podcast hosting platforms for direct selection:

• Pricing model: Single purchase system ($5-20/song) or billing based on usage
• Distribution channel: own website + audio material platform (Audio Jungle, Envato Elements, etc.)

5. Automated workflow upgrade

Use automation tools (such as Zapier, Make) to build a fully automated music production line:

Hotspot monitoring → ChatGPT automatically generates lyrics → Suno API batch generation → Automatic audio download → CapCut automatically synthesizes videos → Releases short videos at regular intervals → Data is automatically written back to the dashboard

Technical threshold: A certain API programming foundation is required, or visual configuration using a codeless automation platform is required.

6. Copyright assetization and NFT exploration

Put original AI music works on the chain in the form of digital assets (such as Ethereum, Polygon), and sell them as NFT music works or share the copyright: single limited edition, series theme EP, real-time generated music NFT (different arrangements are generated each time it is played).