Start with two numbers. In early 2025, Douyin’s monthly top-5000 short dramas contained exactly 4 AI comic dramas; by year’s end, 217. In January 2026, 14,634 AI-generated short dramas (AI live-action plus comic dramas) went live in China — 470 a day, one new title every 90 seconds, per DataEye. Now a third number: in April 2026, roughly 44,000 new AI titles launched in a single month — and 267 of them broke a hundred million plays. A 0.6% hit rate.
Hold those three numbers together and you have the complete mid-2026 truth of this track: the demand is real, the money is real, and the “anyone can cash in” phase is over. This series is for people entering seriously: from production pipeline to platform revenue share, dismantled part by part into executable steps. Part 1 is the map — where the money is, how much window is left, and whether one person can actually pull this off.
What an AI comic drama is, and why it works
An AI comic drama is the fusion of AI + comics + short drama: AI-generated visuals, voice, and motion, telling serialized stories at short-drama pacing (1–3 minute episodes, hard hooks, dense plot twists). Its foundation is a collapse in the cost structure:
- Traditional animation was “a luxury of both time and money” — a minute of quality animation could cost hundreds of thousands of yuan;
- In 2024, AI short-drama production still ran ¥15,000/minute — two to three times the cost of live-action;
- Since 2025, as video foundation models industrialized (Kling, Jimeng, Seedance and their cohort), per-minute cost fell to about ¥1,000, with top studios pushing per-episode cost under ¥500; AI covers roughly 80% of the pipeline;
- One person producing a 3-minute dialogue-driven episode end-to-end takes 3–8 hours — a figure that recurs across multiple teams’ published SOPs.
In one sentence: animation’s expressiveness + short drama’s business model + AI’s cost curve — three things that arrived simultaneously in 2025–2026. On the demand side, the most direct evidence: Hongguo’s free comic-drama app reached 24 million monthly active users within four months of launch.
Where the money is: five revenue models, barriers and ceilings
Money in this track comes in five forms, ordered by barrier:
- Novel referrals — lowest barrier (~1k followers): recreate famous scenes from web novels as comic dramas, cut at the cliffhanger, direct viewers to search a keyword; ¥5–12 per acquired user, weekly settlement. Ideal cash flow during cold start.
- Custom production — no followers needed, portfolio required: countless web-novel studios and short-drama companies hold IP and scripts but lack the AI stack; market rate is ¥300–1,500 per finished minute. B2B cash flow, and the best training ground.
- Platform revenue share — the highest ceiling: Douyin/Hongguo’s “guaranteed purchase + coefficient split,” Tencent’s Huolong (opened to individuals in February 2026, exclusive-content coefficient up to 200%), iQiyi exclusive splits up to 100%. This is the main battlefield for content makers.
- Brand deals — from 10k followers up; vertical accounts quote roughly ¥150–300 per 10k followers per deal.
- IP merchandise — a 100k-follower game with 60–70% gross margins, and print-on-demand means no inventory risk.
How to read the map: the first two feed cash flow, the third builds assets, the last two are amplifiers. This series’ main line is the third, but the first two play key roles in Part 7’s ROI teardown.
Reading the window honestly: first-half data vs second-half data
Is it too late to enter? Split the data into halves and the answer gets honest:
First-half signals (demand and dividends): the market ran about ¥16.8B in 2025 and projects toward ¥24B in 2026; after Hongguo’s short-drama app broke 100M DAU, comic dramas became the incremental category platforms fight over; monthly revenue-share for AI short dramas on Douyin is already at the hundreds-of-millions level (per Ocean Engine monitoring); toolchain maturity is still improving monthly.
Second-half signals (supply and shakeout): 470 new titles a day against a 0.6% hit rate; on April 30, Douyin cut revenue coefficients (photo-real 60→40, 3D 50→40); guarantee prices are falling and “zero-guarantee contracts” have appeared; platforms keep tightening enforcement against reposting, content-laundering, and low-quality AI output.
Only the two sentences together form the complete judgment: the brute-force era of platform-subsidized volume play is over; the content-quality era has just begun. The window isn’t closed — it changed the price of admission. It used to be “make it fast”; now it’s “make it good.” For people willing to actually learn the pipeline, that’s good news: the tools are public; taste and throughput are the moat.
The platform landscape at a glance (Part 6 goes deep)
- Douyin system (Hongguo): the main battlefield — biggest traffic, most frequent rule changes; individual creators onboard through Douyin’s short-drama copyright center by signing the motion-comic licensing agreement.
- Tencent system (Huolong + Tencent Video): opened to individuals in February 2026, aggressive exclusive splits, backed by the China Literature IP library.
- Bilibili: native ACGN soil, suited for 10–30 minute story-driven long-form comic dramas; creator incentives, tipping, and brand deals.
- Kuaishou / RED / WeChat Channels: trailers, serialization, persona accounts — traffic incentives and brand deals.
- Going overseas (TikTok / ReelShort): another 2026–2027 window; the production capability transfers directly.
Can one person do it: pipeline feasibility
“One-person production” is a contested claim in the industry — one team flatly said “impossible,” but their denial came with a suffix: “unless time is unlimited.” Translated: solo is feasible, provided the process is engineered and the genre matches your throughput. The validated solo pipeline has five stages:
Script (LLM-assisted, ~2h) → storyboard table (~1h) → image generation (character consistency is the core problem, ~2h) → motion and voice (image-to-video + TTS, ~1.5h) → edit and publish (templated, ~1.5h). About 8 hours per episode — one to two episodes a week, sustainable for a solo operator. Parts 2–5 of this series turn this diagram into an executable operations manual, stage by stage.
One judgment to state early: every stage of the pipeline can be automated except one — whether this story is worth making at all. A 0.6% hit rate means topic selection and script outweigh production by a wide margin. That’s why Part 3 (script and storyboard) will be the longest part of the series.
Action items, then Part 2
- Build taste by watching: on Hongguo’s comic-drama app and Bilibili, find 5 recent AI comic dramas that broke ten million plays; for each, write three lines — genre, the hook in the first 10 seconds of episode one, and what the comment section is arguing about. That’s lesson one of topic selection, and it’s free.
- Do your own math: using the pipeline’s time budget, compute how many episodes per week you can sustainably ship; then pick your combination of “cash-flow model + main battlefield” from the five revenue streams. Write it down — Part 7 will tear down your ROI against it.
- Register an account: walk through the individual-creator onboarding at Douyin’s short-drama copyright center (publish nothing), turning “I’ve heard of the rules” into “I’ve read the rules” — especially the copyright clauses in the licensing agreement.
Part 2 turns the pipeline diagram into a shopping list: a tool-selection matrix for each of the five stages (free-tier starter kit vs advanced kit), the comparison dimensions across the four layers — text-to-image, image-to-video, TTS, editing — and the minimum viable toolchain a ¥500/month budget can buy.