Published: October 27, 2026

How One AI Rap Video Reached 120 Million Views: 9 Lessons for AI Music Creators

Bottom line: “Guo Lai Ren” did not reportedly reach more than 120 million views because it looked AI-generated. It spread because one creator used AI production speed to deliver a deeply human idea about rent, layoffs, grief, late starts, and perseverance.

That distinction matters for every ai music creator trying to grow an audience.

Better tools can lower production costs.

They cannot manufacture a reason for someone to watch, comment, remember, and share.

What is “Guo Lai Ren”?

“Guo Lai Ren,” translated by the South China Morning Post as “The One Who’s Been There,” is an AI-generated Chinese rap video featuring ten figures from Chinese history and mythology.

The video places figures including poets Du Fu and Su Shi, Han dynasty founder Liu Bang, and the Monkey King in a modern rap performance.

They do not merely recite historical quotations.

They connect their own hardships to present-day problems such as moving house in middle age, employment insecurity after 35, family loss, bills, and the fear of starting too late.

The South China Morning Post reported on October 26 that the video passed 100 million views across platforms in less than two weeks.

A detailed October 23 interview with creator Wang Kunzhe in 21st Century Business Herald put the combined total above 120 million by 9 p.m. on October 22.

That report attributed approximately 75 million views to Douyin, 45 million to WeChat Channels, 3 million to Xiaohongshu, 500,000 to Bilibili, and 150,000 to Weibo.

Those are reported platform totals rather than an independently audited count, but the scale and speed are still notable.

Sources: South China Morning Post, October 26, 2026; 21st Century Business Herald, October 23, 2026.

Who created the viral AI rap video?

Wang Kunzhe, a Wuhan-based creator who publishes as ZheFILM, made the video independently.

He previously spent more than ten years working in television news, promotional films, and commercials before moving fully into AI production in 2026.

His interview says the production itself took three to four days and cost several thousand yuan in subscriptions, memberships, and computing credits.

Yet the finished work was not a three-day idea.

The lyric foundation came partly from a personal notes archive Wang had maintained for at least ten years.

The characters came from seven earlier installments of his “historical figures take the mic” series.

His filmmaking judgment came from more than a decade behind cameras and in edit rooms.

The real production timeline was three to four days plus years of preparation.

Nine marketing lessons from 120 million reported views

1. Lead with a human tension, not the AI tool

The promotional hook was not “look what this model generated.”

The hook was that people remembered for greatness also struggled through instability, rejection, loss, and uncertainty.

That premise gave viewers an emotional reason to stay.

Before generating anything, write one sentence that answers: “Why would this song matter to someone today?”

If the answer is only that the technology is impressive, the concept needs more work.

2. Translate a specific struggle into a universal feeling

Specificity made the story believable.

Du Fu addresses the strain of moving and renting in middle age.

Su Shi embodies career disruption and personal loss.

Liu Bang represents the person who starts late but refuses to leave the game.

The details are culturally specific, but the feelings travel: instability, disappointment, endurance, and hope.

For your next release, choose one concrete situation instead of a vague emotion.

“Anxiety” is broad.

“Waiting for an overdue payment while rent is due Friday” creates a scene.

3. Use contrast to stop the scroll

Ancient historical figures performing modern rap creates immediate contrast.

The format is easy to describe in a caption and easy to understand in a thumbnail.

Effective contrast can come from era, setting, genre, character, instrumentation, or point of view.

A jazz saxophonist inside a futuristic subway, a 1920s vocalist confronting a modern group chat, or a robot band playing an unplugged breakup song can each create a clear visual question.

The combination should serve the song’s meaning, not merely look strange.

4. Give every character one job

Wang assigned the three main verses to three distinct life situations.

Seven additional figures received only one or two lines each as the video accelerated toward a group finale.

That kept a ten-character idea from becoming ten unrelated biographies.

When building a multi-character AI music video, write a role card for each performer:

• Problem: What does this character carry?
• Message: What single truth does the character deliver?
• Visual cue: How will viewers recognize the role instantly?
• Time limit: How many seconds does the character need?

If two characters perform the same narrative job, combine them.

5. Keep the meaning human-authored

According to Wang, he tested AI-written lyrics but rejected the results because forced rhymes and filler sounds weakened the logic.

He wrote the lyric foundation himself.

AI helped with trial and error in rhyme and arrangement, while he decided what to say, how to order it, and which lines remained.

This is a useful division of labor.

Let AI generate options, alternate structures, rhyme lists, and production variations.

Keep the central experience, opinion, and final editorial judgment in human hands.

6. Build reusable character assets before the breakout

“Guo Lai Ren” was the latest episode in an existing series.

Wang reused character images developed during seven earlier installments, which helped maintain visual consistency and reduce production time.

An ai music creator can apply the same system without using historical figures.

Create a character bible containing front, side, and three-quarter views; expressions; wardrobe; color palette; lighting; scale; approved prompts; negative prompts; and reference frames.

Store reusable environments, logos, type treatments, transitions, and end cards alongside it.

The goal is not to repeat the same video.

The goal is to stop rebuilding your visual identity from zero.

7. Treat earlier posts as research, not failures

Wang reviewed audience profiles and comments from previous episodes before making the breakout video.

His earlier installments reportedly attracted mostly men aged 31 to 40.

The new video broadened to a more balanced gender split and expanded in both age directions.

After every release, record:

• the first three seconds with the strongest retention;
• the lines people quote in comments;
• the audience problem that produces personal stories;
• the characters viewers want to see again;
• the point where viewers stop watching; and
• the platforms where sharing outperforms passive views.

Use those signals to sharpen the next concept.

Do not let metrics replace your point of view.

8. Design the comments section into the concept

Reports describe the comment section as a “tree hollow”—a place where people shared their own experiences and comforted one another.

That did more than add engagement.

It turned a finished video into a living community conversation.

End your caption with a prompt that invites experience rather than praise.

Ask “Which verse felt like your year?” instead of “Do you like my new song?”

Respond to early comments with care.

Pin a comment that models the depth and tone you want others to follow.

9. Protect audience trust before chasing revenue

Wang said platform incentives, a creator-growth program, and compute reimbursements were current sources of income.

Brand opportunities arrived after the video became popular.

He also said he planned to evaluate sponsorships slowly because a disappointing commercial post could spend the trust built with viewers.

That is the correct sequence for creator monetization:

1. earn attention with a meaningful work;
2. convert attention into community;
3. identify offers that fit the community; and
4. reject deals that weaken the reason people followed.

Potential revenue can include platform creator programs, video-production commissions, tool partnerships, carefully matched sponsorships, direct fan support, sync opportunities, educational products, and branded storytelling.

Availability and eligibility vary by country and platform, so verify current terms before building a forecast.

A seven-day AI music video sprint

Day 1: Find the emotional sentence

Write the audience problem and the promise of the song in one sentence.

Collect real phrases from your notes, conversations, and comments.

Day 2: Build the narrative spine

Assign one job to each verse or character.

Plan escalation from the opening problem to the final emotional release.

Day 3: Write and edit the lyric base

Draft the meaning yourself.

Use AI for alternatives, not authority.

Read every line aloud and remove phrases that exist only to complete a rhyme.

Day 4: Generate musical versions

Create controlled variations in tempo, instrumentation, intensity, and vocal delivery.

Choose the version that advances the story, not merely the one with the cleanest surface.

Day 5: Produce consistent visuals

Use approved character references and shot rules.

Generate the opening frame, emotional close-ups, transitions, and finale before filling every intermediate shot.

Day 6: Edit for mobile attention

Make the premise legible without sound in the opening seconds.

Add accurate captions.

Change visual information when the lyric’s emotional function changes.

Day 7: Publish, converse, and document

Adapt captions and formatting to each platform.

Invite a personal response.

Save prompts, drafts, stems, source assets, edit files, licenses, and publishing dates as evidence of your process.

Frequently asked questions

Did AI write the lyrics to “Guo Lai Ren”?

Wang said he wrote the lyric foundation and used AI for rhyme and arrangement experiments.

He rejected fully AI-written attempts because they did not preserve the logic and quality he wanted.

How long did the video take to make?

The reported production time was three to four days, or under 100 hours.

That figure does not include Wang’s decade of professional experience, long-running notes habit, or assets from the seven earlier videos in the series.

Which AI music tool did the creator use?

The detailed interview names LibTV as the main visual workspace but does not identify the AI music tools used for the track.

The strategic lesson is therefore tool-independent.

Should creators copy the historical-character format?

No.

Copy the underlying principles: a clear contrast, specific human tension, disciplined structure, reusable assets, strong editorial judgment, and a concept that invites community response.

Can the 120 million figure be treated as audited reach?

No.

It is an aggregate reported by Chinese media from multiple platform counts.

Platform totals may use different counting methods and should not be treated as independently audited unique viewers.

The real lesson for AI music marketing

AI made this ambitious production possible for one person in a few days.

It did not supply the decade of notes, the storytelling instincts, the earlier episodes, or the emotional truth.

For your next release, spend less time asking whether the output looks advanced.

Ask whether the idea gives viewers somewhere to place their own experience.

That is where sharing begins.

Sources

• 21st Century Business Herald: Interview with creator Wang Kunzhe (October 23, 2026)
• South China Morning Post: Why China’s viral AI rap track struck a chord (October 26, 2026)
• Xinhua: “Guo Lai Ren” exceeds 124 million reported views (October 26, 2026)
• The AI Musicpreneur: English-language report and source summary (October 26, 2026)