Making Money from AI Music Is Far Harder Than Cashing In on Short Dramas

11/20 2026 378

Source | SourceSight

In the first half of 2026, more than 1,500 AI-generated short dramas flooded the market every day. Over the same period, the market also received at least 75,000 AI-generated songs daily—a figure that had climbed to 90,000 by September.

These numbers come from Deezer, the French music streaming platform. In January 2025, AI songs accounted for just 10% of its daily new uploads; today, that share is approaching 50%.

This massive supply stems from a rapid surge in productivity. Alibaba, NetEase, Tencent, ByteDance, and a host of vertical AI companies have all rolled out AI music models or related agents. Unlike AI short dramas, which have already settled into a recognizable market landscape, China's AI music sector has yet to see a single dominant player—which is precisely why Big Tech remains so eager to invest.

On the flip side, AI short dramas have captured a sizable audience, while AI music struggles to find listeners. According to Deezer's data, these AI tracks—which make up 50% of the platform's catalog—garner only 1% to 3% of total play counts.

On November 17, Tencent Music Entertainment Group released its Q3 financial report. On the earnings call, management noted that streaming play counts for AI-generated content "remain negligible."

How much real value is there in cultural and artistic works generated with a single click? The data speaks volumes. Even after "creative equality" has been achieved, audiences' choices remain remarkably limited. Having secured users, the AI music track will likely still need to keep searching for real listeners.

01

Who Exactly Is Listening to AI Music?

While the AI short drama Armed Into Song was trending, the AI song "Come Before" also took off. This rap number—performed by a chorus of 10 ancient figures—became another AI music phenomenon that sparked widespread social discussion, following last year's "Big Head Pin."

In an interview, creator Wang Kunzhe explained that he fed an earlier demo into an AI music tool. After referencing it, the AI generated an entirely new piece of music. The lyrics were written by the author and refined with AI, and an MV was ultimately produced using an AI video tool. To date, the MV has racked up over 100 million views across the internet.

Screenshot from Douyin

On closer analysis, the MV probably deserves much of the credit for "Come Before" going viral. Had the song appeared only on music streaming platforms, it likely would never have attracted such attention.

A US entertainment consumption survey conducted this spring by music industry data agency Luminate found that "when US listeners know a song is AI-generated, their willingness to listen drops sharply into negative territory." This negative attitude is spreading rapidly among younger demographics. The survey results show that AI music faces a major challenge in earning emotional recognition.

In the film and television content market, better models combined with higher-cost computing power—producing more spectacular visuals and more imaginative "spectacles"—have allowed AI short dramas to quickly win over audiences. But when the same logic is applied to music, many people feel a pang of disappointment upon discovering that a song they like was made by AI.

Screenshot from Xiaohongshu

On domestic social media, people discuss how to make money with AI music and how to get the most out of AI music tools, but relatively few recommend actual AI music works.

To some extent, this has shifted the capital market's attitude toward AI music.

The market once believed that a flood of cheap AI cover songs would upend the existing streaming subscription model—to the point where music copyright would no longer serve as a platform's moat. Current trends, however, suggest that the threat posed by AI-generated music has diminished considerably.

On Tencent Music's Q3 earnings call, management noted that technological advances have made content supply more abundant, but play counts for AI content are negligible—and high-quality, differentiated content is drawing greater attention.

This has also narrowed the current AI music race to one goal: capturing users' attention. Whoever can first get users to "play" with AI music in a simpler, more approachable way wins the first victory.

02

Who Will Win the Chinese Version of "Suno"?

Yet the public's vigorous creative appetite still needs to be satisfied—and Big Tech has been quite active on this front.

Alibaba recently launched the "Happy Shrimp" app. Half of the name continues the "Happy Family" branding of Alibaba's ATH business group; the other half borrows from the now-defunct Xiami Music.

As early as August, the "Happy Shrimp" AI music model had already been released. It joins Alibaba's growing lineup of multimodal content creation tools, following HappyHorse (focused on video generation) and HappyOyster (focused on world models).

Opening the Happy Shrimp app, you'll find a lightweight AI music creation tool. The homepage features a recommendation feed where you can swipe to switch songs; for listening, the app focuses on appreciating AI music shared by other creators. The core function, though, is still "write a song with one sentence." A simple Chinese pop song can be generated in under 5 minutes for 20 points—which, converted to membership pricing, amounts to just two mao.

Image source: Happy Shrimp

Beyond Happy Shrimp, virtually every major internet company has staked out territory in AI music. Tencent launched "Youyin" last year and this year rolled out the intelligent creation platform "MusicBuddy" in a race-to-market fashion. The platform is currently still in internal testing.

ByteDance has deployed the Haimian Music app. This product is completely free—users earn creation points by logging in and completing sharing tasks. It's powered by ByteDance's self-developed SeedMusic model. ByteDance also launched the Qishui AI Music Creation Lab. NetEase, for its part, released NetEase Tianyin.

Since 2023, the global AI music track has continued to heat up, producing an industry-level benchmark application: Suno. Its valuation had reached $5.4 billion by June of this year, with paying users exceeding 2 million.

Even though domestic Big Tech companies have been following one after another, no clear leader has emerged. Instead, some vertical third-party platforms have recently turned in notable performances.

Recently, the music vocal ranking released by globally authoritative AI evaluation agency Artificial Analysis placed a Chinese team—ACE Studio—at number one. Its ACE Music v1 achieved a score of 1160, higher than Suno v6.

ACE Studio founder Guo Jing recently told LatePost in an interview that the company's annual revenue has reached $20 million, with paying users worldwide, primarily in European and American markets.

According to reports, ACE Studio has completed two rounds of financing totaling more than $50 million, making it the most heavily funded team in China's AI music sector at present.

From mass-market, entertainment-oriented consumer apps to professional AI software for working musicians, China's AI music landscape has already gathered a considerable number of products. The "big companies" want attention and traffic and need to stake their claims early; the "small companies" are taking the technology route, hoping to secure higher valuations and more valuable professional users.

Having solved the creation problem for AI music, the next challenge is distribution and revenue balance across the industry chain.

03

Copyright Holders' Compromise with AI

Even when facing audiences with mixed attitudes, AI music development can still learn from AI short dramas.

At the outset, platforms attract creators through subsidies and traffic recommendations, increasing the odds of producing hit short dramas—which in turn draws advertisers and viewers, completing the revenue cycle. The same playbook applies to AI music.

Alibaba's Happy Shrimp chose Taihe Music as a partner; the two sides will explore directions such as co-building the AI music industry ecosystem and musician co-creation.

Tencent Music's AI music intelligent creation platform offers "one-click distribution," covering hundreds of channels including QQ Music, Kugou, and Kuwo; NetEase and Douyin have corresponding platform resources as well.

Each has also established its own AI creation incentive programs. Douyin, for example, launched the AI New Voice Plan, while NetEase has held AI music creation competitions for two consecutive years.

Although AI music's overall play counts remain far from mainstream, occasional hits do emerge. Even many ordinary people with no musical background have relied on AI to upload a large number of works, gaining play counts and earning subsidy income from platforms.

Last year, the AI song "Seven-Day Lover" went viral, surpassing 2 million plays in 4 days. In the end, however, the royalties for the song sold for only 50,000 yuan.

Compared with star singers at the same playback level, AI songs are simply too cheap.

The copyrights of top singers are core IP assets, but the massive volume of AI music beneath the surface also holds the possibility of growing into IP.

Tencent Music once again emphasized the importance of IP on its latest earnings call. Q3 total revenue was 8.94 billion yuan, up 5.6% year on year, of which membership service revenue was 4.99 billion yuan (up 9.2% year on year); marketing and consumer services revenue was 2.70 billion yuan (up 12.1% year on year), driven by artist merchandise sales and offline performance-related services.

Meanwhile, foreign streaming platforms, under the current copyright landscape, have already imposed restrictions on recommendation and distribution procedures for AI music.

Warner Music reached a settlement with Suno, and Universal Music reached a settlement with AI company Audio. Multiple parties have formed a community of interests. Music generated by unauthorized AI training will become the next target of copyright holders and platforms. Under these circumstances, platforms that control ecological entry points and traffic also need to establish "legalized" AI music libraries.

AI has reshaped the music production model, and lowering the barriers to creation has value in itself. But the market previously overestimated AI's ability to replace high-quality, differentiated human music in the short term, while underestimating the long-term costs of platform governance and cultivating listeners' aesthetics.

Some images are sourced from the internet. If there is any infringement, please inform us for deletion.

Solemnly declare: the copyright of this article belongs to the original author. The reprinted article is only for the purpose of spreading more information. If the author's information is marked incorrectly, please contact us immediately to modify or delete it. Thank you.