AI comic drama quotes fall from 5,000 yuan/minute to just tens of yuan; one company can produce 600 episodes a day

"I lost an AI comic-drama order worth over 100 yuan per minute, and I was so upset I couldn't sleep..."

Recently, a public complaint by an AI short-drama blogger on social media struck a chord with many people in the industry. In the person's view, this was a missed money-making opportunity. But in the eyes of peers, the price anxiety in the AI comic-drama circle has now reached a point where it needs to stop—from 5,000 yuan per minute at the start of the year to 200 yuan per minute now, and even just a few dozen yuan per minute... You should know that many AI comic-dramas often have only one minute per episode, and such a sharp price decline directly outlines the industry's "unprecedentedly competitive" situation.

In conversations with the "BUG" column, several veteran industry figures admitted that the decline in AI comic-drama prices had shown signs early on. The rapid progress of AI technology is causing the industry's production capacity to explode at a rate of more than 10 times, and the corresponding production quotes are also falling to one-tenth of the original or lower.

What is even more brutal is that, because the overall audience for AI comic-dramas currently remains at around 100 million people, the unprecedentedly explosive production capacity of AI comic-dramas is also diluting the probability of quality comic-dramas being seen to one-tenth of the original or even lower. Without professional agencies handling promotion and traffic buying, it will be much harder for an AI comic-drama to break out.

At present, AI comic-dramas in the commonly understood domestic sense can roughly be divided into three tiers: low-end dynamic comics—which can be understood as animated PPTs; mid-tier premium comic-dramas; and high-end 3D modeling comic-dramas. The production quotes for these three tiers rise from low to high, but all have seen significant declines over the past year, with the low-end track being the most fiercely competitive.

According to Peng Xiang, an AI comic-drama creator and head of the brand Maopai Xingxing, who spoke to the "BUG" column, the cost of low-end dynamic comics has now fallen to as low as a few hundred yuan per episode. Because they can be mass-produced with AI, this has become the hardest-hit area of the price war, and some creators' quotes have dropped to just a few dozen yuan per minute, while an episode is often just one minute long.

In addition, prices for the second tier, premium comic-dramas, have also dropped to "1,000-2,000 yuan per episode." Because this requires deep human involvement to control the art style and camera movement, with gacha artists, prompt specialists, directors, and multiple manual verification steps, the core is labor cost. Prices for this type of work have also fallen from more than 5,000 yuan at the start of the year to the thousand-yuan range.

The third tier, high-end 3D modeling comic-dramas, has no clear upper cost limit. "The expense mainly lies in labor. The labor quotes for directors and modeling create the gap, not the cost of AI computing power itself." Such customized works can still maintain a relatively high premium, but with the improvement in AI modeling capabilities brought by the release of GPT-6, this premium has also declined noticeably.

The core reason for the price collapse is the explosion of production capacity on the supply side. AI has enabled assembly-line mass production. Low-end dynamic comics only require purchasing web-novel IP and inputting it into a workflow to automatically split shots, generate images, and convert them into video, with very little human involvement. Peng Xiang admitted, "I know of one comic-drama company that, at its extreme, could produce 600 episodes of comic-dramas in a single day. That's too exaggerated!"

A large number of newcomer teams have poured into the track, constantly undercutting quotes to grab orders, directly driving industry prices down by dozens or even a hundred times.

"Make tens of millions with a small budget? That's all survivorship bias"

According to media reports, in the first eight months of this year, 430,000 micro-short dramas were launched domestically, 13 times the total for all of last year, of which AI dramas accounted for more than 90%, mainly because artificial intelligence technology has been widely applied in micro-short drama creation.

AI technology has opened up new space for micro-short dramas to reduce costs, improve efficiency, and innovate expression, while also bringing new challenges such as rights protection and content quality. With the supply explosion and severe involution, after an AI comic-drama is made, doing a good job of traffic exposure has also become a field of expertise. In conversations with the "BUG" column, Chen Shaofeng, vice dean of the Institute for Cultural Industries at Peking University, admitted that AI short dramas already "have no future," but AI short dramas + brand communication are becoming a "big business."

When paid short dramas were booming in earlier years, there was a saying in the industry: "10% production, 90% traffic buying." As a veteran with nearly ten years in the industry whose team's live-action short-drama annual production capacity reaches 2,000-3,000 episodes, Wang Xukun, founder of Xiaoxiang Yinghua, also admitted: "For teams that only do pure production and do not control distribution, profits are actually very thin."

English text is being prepared.

He further revealed that with the substantial increase in short-drama supply, the thinking of top short-drama and comic-drama players has shifted from a "hit-driven mindset" to a small-step, fast-running mindset of "stable supply and protecting principal": "My first priority in making short dramas is to protect principal, to get my money back in the shortest possible time. I never pursue making a hit. Making a hit is just for fame, but I treat this as a business."

Wang Xukun believes that the core advantage of top teams is having stable production capacity supply and work quality assurance, which makes platforms willing to rate scripts before short dramas even start shooting or are officially approved, and calculate the return on investment in advance. But not everyone has this qualification: "80% of teams cannot do this."

The result is that the vast majority of small and medium-sized comic-drama/short-drama teams follow the path of "find a script, scrape together money, produce it, then sell the finished product," like "selling used cars," where "the longer you wait, the more it depreciates." In essence, it is gambling on probability.

“The stories circulating online about making tens of millions on a shoestring are all survivorship bias.” Qi Yu, a veteran micro-short-drama industry practitioner, was equally blunt: platforms will not bet on “the next hit director,” and the works of a large number of small and medium-sized teams will ultimately sink to the bottom of the content pool, ignored by all.

“Trainees turned into cheap card-drawers”

In conversations with the 《BUG》 column, several industry insiders admitted frankly, “The vast majority of AI comic-drama training on the market is heavily watered down; AI training courses are basically scams.” The view of frontline creators is very direct: “This stuff you can just study on your own—do you even need training? It’s like learning to ride a bicycle; it depends on personal aptitude.” The operational threshold for AI tools is extremely low, and the basic workflow can be mastered through self-study. Training courses mostly teach only the surface, and the true core—aesthetic judgment and content judgment—simply cannot be taught through short-term training.

Even more alarming is the emerging “training + factory” scheme. Many training institutions themselves own short-drama companies: they collect tuition of around 20,000 yuan while converting trainees into cheap “card-drawers” (AI video creators), drawing cards for their own projects under the guise of practical training.

“They need a large number of card-drawers; it doesn’t require much technical skill. Once you understand the workflow, you become a cog in their machine,” one practitioner admitted frankly. The real purpose of many institutions’ recruitment is to screen for low-cost assembly-line production staff. Trainees pay the money and instead become the other side’s production tools.

Faced with the industry reality of price involution and exploding production capacity, the advice from interviewees is highly consistent: don’t get trapped in a price war; build genuine content barriers. “Don’t do every subject—like a chef, being good at Sichuan cuisine doesn’t mean you’re good at Cantonese cuisine.” Wang Xukun advises newcomers to firmly identify their track anchor: “Make the track narrow and deep—for example, specialize in revenge comic dramas about middle-aged women, and thoroughly understand the corresponding audience, aesthetics, and topics.” Don’t fantasize about dominating every genre.

Second, don’t fetishize tools as omnipotent. “AI can only create short-term sensory stimulation; genuine emotion still depends on people.” As a top micro-short-drama producer, Wang Xukun admitted frankly that screenwriting skills and insight into human nature and society—these traditional abilities—will never be replaced by AI. This is the true core competitiveness.

Finally, control risk: surviving long matters more than running fast. “Compliance is a moat, not an admission ticket”; ignoring compliance will directly lead to project losses. At the same time, control costs and preserve cash flow: “The industry goes through a generational shift every 3 months; you have to ensure you can survive at least two cycles.”

This article is from the WeChat public account “Sina Tech,” author: Zhou Wenmeng, published by 36Kr with authorization.