Intro: A Hit Whose Formula Got Misread
In the 2026 summer corridor, Baxian! became a phenomenon for Chinese animation: 23 days in, it had crossed 1.3–1.4 billion yuan, with Maoyan/Beacon predicting a final landing around 2.1 billion, and a Douban opening score of 8.3. The line that spread widest was “100 million cost, a billion-plus box office — a tenfold return.” For anyone grinding in the short-form animation (manju) space who occasionally wonders whether to leap into film, that number is almost a temptation.
But mistaking box office for profit, production cost for total cost, and one hit for a repeatable model — get any of the three wrong and you’ll lose badly in this high-leverage industry. This article opens Baxian!’s books and hands a cold-eyed checklist to anyone tempted to jump from short content into feature film.
1. The “AI Traces” You Saw — The Studio Admitted Them
Many viewers left the theater feeling the film was “too AI”: the protagonist’s face drifted between shots, some segments switched styles abruptly, the lighting had an indescribable “greasiness.” This isn’t audience paranoia.
The end credits literally list “AI Creative Production: Wuyue Chuangxiang” and a “AI Creative Production Director” credit — the studio acknowledged a dedicated AI production unit at the credits level. Pearl Studio (Oriental DreamWorks) president Ying Xujun publicly said the team “deeply used AI tools” in art and asset production, and early CCTV coverage called the film’s “AI-assisted VFX rendering” a breakthrough.
What’s telling is that the director later changed his tune, insisting on “full real motion-capture, real production” and that “AI video precision is insufficient for the big screen.” That contradiction itself shows: how much AI participated, and at which stage, is an undisclosed gray zone. The frame-by-frame problems crowdsourced on social media — “greasy lighting, templated expressions, jarring style jumps” (the named example was the “tunnel-to-pigpen” bit) — line up exactly with audience instinct.
Worth noting on the regulatory side: the NRTA’s existing “AI short-drama labeling rule” does not cover theatrical film. So how much AI a theatrical film uses, and how much the studio discloses, is basically voluntary. To judge whether a theatrical animation is “AI-assisted,” audiences mostly still rely on their own eyes — checking style consistency frame by frame, templated expressions, whether lighting holds together across segments. The Baxian! controversy was effectively a free public masterclass in spotting AI-generated film.
2. The Stitched Script: A “De-deified” Rewrite the Whole Internet Panned
There’s a structural controversy on the script level too. Viewers called it a “Frankenstein”: the immortals were secularized wholesale — Lü Dongbin stripped of his celestial rank, Zhongli Quan turned into a street pickpocket, Cao Guojiu written as a corrupt official — a “de-deification” rewrite of the traditional myth. There are even historical howlers: a Ming-dynasty setting mixed with a Qing-era queue hairstyle. On Douban, people compared it to Hollywood ensemble heist comedies like Ocean’s Eleven, arguing the narrative was “stitched” rather than organically grown from folklore.
This has nothing to do with AI, but it reveals another kind of low-cost shortcut: using genre formula to reassemble a public IP, skipping the hard work of original world-building and character arcs. For anyone trying to live on content, this is a “laziness trap” worth flagging — stitching lets you ship fast, but a “opportunistic” ceiling will press down on your reputation.
3. The Maker Isn’t a Small Shop: Pearl Studio + Maoyan
A common guess: Baxian!’s production company wasn’t really disclosed, probably a small firm that outsourced everything, with the 100-million-plus cost full of water. The premise doesn’t hold up.
Baxian! is led by Chengdu Huameng Chengzhen Film & TV Culture Media Co., i.e. Pearl Studio’s Southwest HQ; co-producers include Shanghai Huaren Pictures and Tianjin Maoyan Media. Pearl Studio is a top-tier Chinese animation house (originally the DreamWorks China JV, involved in projects at the Kung Fu Panda 3 scale), and Maoyan is China’s largest ticketing and distribution platform. This isn’t a “small shop, full outsourcing” picture — it’s a major studio plus a top distributor.
That matters: it directly undermines the “the cost is all left-hand-to-right-hand kickbacks” reading.
4. The Real Cost: 117M Production + 150M Marketing
The other simplification is the cost structure. 116M (publicly ~117M yuan) is production cost — animation, VFX, rendering. But a theatrical film’s all-in cost also includes marketing and distribution (promo + release/screen allocation). Per public reporting, Baxian!’s all-in cost (production + marketing) is around 260–280M yuan, meaning marketing burned about 140–160M, borne mostly by Maoyan.
Marketing cost exceeding production cost is the norm in theatrical film today, not an exception. Which is why the next section says marketing is the real deciding factor.
5. The Split Math: The Producer Gets ~33%, Nets 400–500M, Not 10×
Now the key accounting. Box office ≠ producer revenue.
China’s box office is first docked two cuts: the National Film Development Fund (5%) plus VAT and surcharges (~3.3%), about 8.3% off the top. The remaining ~91.7% is the “shareable box office.” That splits again: cinemas and cinema lines take ~57%, the distributor ~4–7%, and the producer/rights holder nets roughly 33% of total box office.
A rough bill at the predicted 2.1B total:
- Producer gross revenue ≈ 2.1B × 33% ≈ 690M
- Minus production cost 117M
- Minus marketing share (marketing is mostly borne by distributor Maoyan, but the producer also carries part)
The producer nets roughly 400–500M. Investing 117M in production to net 400–500M is a ~4–5× return — an undeniable smash hit. But it’s not 10×, and definitely not “100M in, a billion-plus straight into pocket.” The popular “10×” treats box office as the producer’s money, off by 3× in magnitude.
This gap isn’t nitpicking. It decides whether you let that number talk you into jumping ship.
6. Kickbacks / Fake Invoicing Are Real in Film — But Don’t Apply to Baxian, and They’re a Landmine
“Outsourcing quoted high, cash kickback to the producer,” “reporting a 300K job as 2M,” “using cost as a marketing flex” — these practices have existed in China’s film industry, especially during its runaway years. Chinese audiences do treat production cost as a sincerity signal, and “big production” adds marketing value, which gives motive to inflate.
But applying that to Baxian! inverts the logic. 117M production cost isn’t high for a theatrical animated feature — it’s low; top domestic studios (ChLight, Coloroom and peers) generally run theatrical animation above 100M. Its high return comes from real causes: Chengdu’s industrial cluster compressing production cost, AI further cutting cost, genuinely good word-of-mouth, Maoyan’s strong distribution, and well-timed summer release. This is a real hit, not kickback-inflated paper prosperity.
More importantly: don’t romanticize “fake invoicing” as your future film business model. Forging value-added-tax invoices is a crime written into China’s Criminal Law, same category as the helping-information-network crime — a “gray but genuinely prosecutable” landmine. You can keep it as industry-recognition knowledge, but it isn’t profit energy, it’s prison energy.
7. Survivorship Bias: ChLight’s Zheng Luoyang Is the Denominator
The most dangerous thing about “100M for a billion-plus” is survivorship bias. You only counted the numerator — the hit Baxian! — not the denominator: the films that flopped, got pulled from theaters, and went unmentioned.
The same summer corridor offers a ready counterexample: ChLight (the top studio behind White Snake and Nezha) released Zheng Luoyang in the same window, pulled in only 80M-plus, and yanked it from theaters. If a top industrialized studio like ChLight can flop, someone who has never made a film and doesn’t understand distribution will, with near certainty, end up in the denominator, not the numerator.
Hits are visible; flops are buried. For every “this film made dozens of times its cost” story, first ask: how many flopped in the same year and window? The cost of those flops is the industry’s true average.
8. Distribution Is the Moat; Production Is Commoditized
What’s scarce and box-office-decisive in film isn’t production capability — it’s distribution + IP + screen allocation. Baxian!’s explosion owed at least half to Maoyan’s distribution (preview strategy, screen-grabbing, word-of-mouth ops), and marketing cost (150M) exceeded production cost.
Your edge — an AI production pipeline for manju — is a “production cost-reduction” capability. But in film, production is commoditized and outsourceable: anyone can hire a Chengdu render farm, anyone can get an AI-assist pipeline. What’s genuinely scarce is distribution relationships, IP reserves, and screen-allocation leverage. Entering film now means burning your least-advantaged money where you have zero moat.
9. Manju’s Lower Ceiling Is Your Protection, Not a Flaw
Manju’s “energy isn’t as big as film” — the per-title return ceiling is indeed much lower. But low ceiling = small per-unit investment = cheap, fast trial-and-error. Your AI production pipeline is a real cost edge in the manju lane; placed in theatrical film, that edge is diluted to invisibility by 100M-class production + marketing.
The “sports master’s + data background + small team + production pipeline” combination matches manju’s unit economics, not theatrical film’s. Manju’s ceiling is, in essence, locking your downside into a magnitude you can survive.
10. If You Can’t Let It Go: Lean In, Don’t Leap
If you genuinely can’t drop the film idea, there are two low-risk paths — rather than the “raise 100M as an independent producer” bankruptcy move:
- Become a production / asset outsourcing supplier: sell your AI production pipeline to film studios — Pearl Studio itself is deeply using AI; this industry has real demand for “low-cost, high-quality production capability.” Earn supply-chain money, don’t bet on box office, don’t carry marketing.
- Co-production: find a studio with distribution + IP (Maoyan, Enlight, Coloroom) for a joint production — you bring production / AI capability, they bring distribution + release, sharing risk and revenue. Hand your weak spot to a partner with a moat instead of carrying it yourself.
Both paths share one core: turn your thinnest position (production) into others’ most-needed supply, and hand your most-missing position (distribution) to a partner with a moat. That’s “leaning into the film supply chain,” not “leaping into the film casino.”
Closing: Don’t Let Survivorship Bias Push You Off the Ship
On Baxian!’s AI traces and stitched script, your eyes didn’t deceive you — three pieces of evidence (end credits, studio’s public statements, audience frame-by-frame analysis) all confirm it. Keep that “spotting AI-generated film” instinct; it’ll serve you in manju selection and outsourcing acceptance later.
But “100M for a billion-plus” is a denominator-blind illusion. The producer nets about a third of box office; the real return is 4–5× not 10×; all-in cost is 260–280M not 100M-plus; ChLight’s pulled film is the invisible denominator. Manju is the pit you can win now; film is the pit you can’t win until you hold the distribution, IP, and release-relationship cards.
Being hooked into jumping ship by one hit is the most classic opening move for a newcomer in this industry. With a cold-eyed checklist in hand, you won’t be the next denominator.