/logo.webp){kind=logo}
How to Make a Short Drama: Script to Screen
Short dramas — vertical, minute-long episodes that end on a cliffhanger — have gone from a Chinese app curiosity to one of the fastest-growing entertainment formats in the world. Omdia projects the global micro-drama market at $11 billion in 2025, nearly double the size of the entire FAST-channel business. The barrier to entry has never been lower, but the format has hard rules, and most first attempts break them. This guide walks the complete pipeline — format, script, storyboard, characters, production, publishing and season planning — whether you shoot with a crew or generate every frame with AI.
Step 1: Understand the format before you write a word
A short drama (also called a vertical drama or micro drama) is a serialized story told in 9:16 episodes of roughly 60–120 seconds. Every episode follows the same skeleton — instant hook, escalating conflict, cliffhanger ending — and a full title chains 50–100 episodes into one binge arc built for phone feeds.
Think of a title as one movie chopped into micro-chapters: TheWrap describes the standard package as 60 to 100 minutes of total runtime divided into roughly 90-second episodes, with the first 8–10 episodes free and the rest behind per-episode unlocks or a subscription. That business model is not trivia — it dictates the writing. Omdia’s research (via C21Media) found that more than 60% of global micro-drama revenue comes from those subscription and pay-per-episode channels, with per-user spending reaching $20 a week among committed viewers. Viewers pay because a cliffhanger made not paying feel worse.
Three format constraints follow directly. First, the hook must land in the opening three seconds — the scroll is your real competitor. Second, most of the feed watches on mute, so the story must survive with captions alone. Third, every episode ending is a sales pitch for the next one.
Practical rule: Design for the paywall from day one. Most platforms give away the first 8–10 episodes, so the cliffhanger around episode 10 — not episode 1 — is the moment your audience decides whether to pay.
Step 2: Write the script — tropes, beats, one turn per scene
Short drama scripts start from a proven trope, compress it into a beat sheet — hook, conflict, turn, climax, cliffhanger — and break every scene down to a single spoken line per shot. The first three seconds pose a question; the final three seconds open the next episode.
Start inside a lane the audience already binges: CEO romance, contract marriage, rebirth revenge, hidden heir, werewolf mate. This is not creative laziness — it is pattern economics. Viewers arrive with expectations, and your job is to deliver the expected payoff through an unexpected route. From the trope, build 6–10 beats per episode with exactly one reversal per beat, and place a mini-cliffhanger around the 15-second mark, the first big drop-off point on short-form feeds.
Writing in shots matters more than writing pretty prose. A line of dialogue that cannot be captioned in one screen is too long; a scene that needs a paragraph of setup is two scenes. An AI script generator built for this format writes that way natively — beat-timed scenes, one line per shot with emotion cues, and a six-dimension review (hook speed, information density, conflict, character clarity, cliffhanger, platform compliance) that scores each draft out of 100 before you commit any production budget. You can also paste a viral drama link or transcript and get its beat structure and emotion curve back as a reusable template, plus two or three alternative openings to A/B test.
Practical rule: One turn per beat. If a scene contains two reversals, split it into two shots; if it contains none, cut it. Flat middles are where short dramas die.
Step 3: Storyboard the episode shot by shot
A storyboard converts your beat sheet into a numbered shot list — framing, camera move, caption space — before any money or render time is spent. Because a short drama episode is only 6–10 shots, a whole season’s board is a few hundred deliberate frames, not thousands, which makes this the highest-leverage planning step in the pipeline.
Vertical framing is its own craft. The Los Angeles Film School’s look at ReelShort-style production notes that techniques like rack focus thrive in 9:16 while wide, background-heavy moves such as the dolly zoom lose their impact in a narrow frame. In practice: faces dominate, two-shots get tight, and the lower third of every frame belongs to captions.
If you are new to boarding, start with the fundamentals in our guide to how to storyboard a short film, then go deeper where short dramas get hard: storyboarding action scenes for confrontation and chase beats, and the Sora 2 storyboard workflow if you plan to drive AI video generation from your boards.
Practical rule: Board in 9:16 from the first frame. A board drawn widescreen and cropped later loses exactly the edges where vertical storytelling lives — faces, hands and caption space.
Step 4: Lock characters and continuity before you produce anything
Continuity — the same face, wardrobe and visual logic across every episode — is the most fragile thing in short drama production, because one title means hundreds of scenes with the same leads. Lock a character bible before producing a single scene, and treat any change to it as a production decision, not a whim.
The scale is the problem: The Hollywood Reporter puts a typical vertical series at 50 to 80 episodes, shot at a pace of three to five takes per setup. On a traditional shoot that means wardrobe bibles, continuity photos and a script supervisor working flat out. In AI production the same failure mode appears as character drift — a lead whose face subtly changes between shots, which breaks the binge instantly. Purpose-built tools solve it with character lock: one portrait per character anchors face, wardrobe and vibe across every scene, and each new episode inherits the cast, voices and visual language of the last, so your episode-40 heroine is recognizably your episode-1 heroine.
Practical rule: Cast once, reuse forever. Whether your actors are human or generated, every re-cast mid-season costs you the viewers who came back for a face they already trust.
Step 5: Produce the scenes — traditional shoot vs AI generation
There are two honest paths from storyboard to footage: a traditional vertical shoot with real actors, or AI generation that renders your board into finished, dialogue-synced scenes. They differ by orders of magnitude in cost and speed, and they are good at different things — many studios now use both.
The traditional numbers are well documented. TheWrap reports typical budgets of $150,000 to $300,000 per title on an 8-to-10-day production cycle; The Hollywood Reporter quotes ReelShort’s CEO keeping projects under $300,000, with a feature-length script often shot in about a week of 12-hour days. Before the camera rolls, Casting Networks’ guide to verticals describes 3–4 weeks of pre-production and casting, with day rates of $100–$500 for most roles.
| Traditional shoot | AI generation | |
|---|---|---|
| Budget per title | $150k–$300k | Software credits; free to start |
| Timeline | 3–4 weeks prep + 8–10 shoot days | An episode in an afternoon |
| Retakes | 3–5 takes per setup | Unlimited; every take kept, roll back anytime |
| Crew | Full cast and crew | One person |
| Strongest at | Nuanced human performance, tactile realism | Speed, iteration, testing hooks before green-light |
What the shoot still wins: the micro-expressions of a good actor, real chemistry between leads, and the texture of physical locations. What AI wins: cost, speed, and iteration — an AI drama generator takes a script through character portraits, multi-voice dubbing, word-level lip-sync and storyboards to a captioned 9:16 export in one place, and a bad take costs a regeneration instead of a reshoot. For stylized or animated looks, see how to turn a storyboard into animation. The emerging hybrid play: prototype ten hooks with AI, read the data, then green-light the proven winner for a full shoot.
Practical rule: Prototype with AI even if you plan to shoot. Testing ten hooks in software costs less than one hour of a ten-day set — and the retention data tells you which script deserves the $200k.
Step 6: Edit, caption and publish vertical-first
Post-production for a short drama means ruthless cuts, burned-in captions and a native 9:16 export — never a widescreen video letterboxed into a feed. The edit’s only job is momentum: anything that does not advance the conflict or set up the cliffhanger gets cut, no matter how good it looks.
Captions are not an accessibility extra here; they are the primary dialogue channel for the majority who scroll on mute. Burn them in, keep them inside the platform safe zone (clear of the like button and the caption bar), and time them line by line so the twist lands with the cut. Export a captioned 9:16 MP4 sized for TikTok, Reels, Shorts and WeChat Channels, then publish while the idea is fresh and read the retention curve — the second-by-second drop-off graph is your writers’ room for episode two. For platform-specific hook placement, posting cadence and caption tactics, see the companion guide: how to make an AI short drama for TikTok.
Step 7: Plan the season, not the episode
A single episode is a test; the season is the product. Before rendering episode two, map the season arc — premise, per-episode hooks, and a cliffhanger chain in which every ending bites into the next opening — so binge momentum is designed in rather than hoped for.
Velocity is the moat in this market. TheWrap reports ReelShort’s team alone turning out 15–20 vertical scripts a month — a pace no one matches by treating each episode as a fresh creative project. Systematize instead: a season planner that grows one premise into up to 60 episodes with hooks and cliffhangers mapped up front, and episodes that inherit your cast, voices and look automatically, leave you making one real decision per episode — is this twist strong enough? — instead of fifty. Close the loop with data: if episode one’s retention dips at second 12, the fix belongs in every future hook, not just the next one.
FAQ
How much does it cost to make a short drama?
A traditionally shot vertical drama typically costs $150,000–$300,000 per title, per TheWrap and The Hollywood Reporter’s industry reporting. An AI-generated short drama costs software credits — DramaSo is free to start, with welcome credits that cover a scripted, dubbed, captioned first episode.
How long does it take to make a short drama?
Traditional production runs 3–4 weeks of pre-production plus 8–10 shoot days of up to 12 hours, then editing. With an AI pipeline, a solo creator can take one episode from logline to captioned 9:16 export in an afternoon, and a season in weeks.
How many episodes should a short drama have?
Industry-standard titles run 50–80 episodes of 1–3 minutes each. Plan at least a full arc of 60, front-load your strongest hooks in the free episodes (usually the first 8–10), and put a season-defining cliffhanger right at the paywall.
Can I make a short drama alone, without a crew?
Yes. An end-to-end AI studio covers the whole pipeline — script with a quality score gate, storyboards, locked characters, multi-voice dubbing with lip-sync, and captioned vertical export — so one person can run what used to need a set.
What format should a short drama be?
Vertical 9:16, episodes of roughly 60–120 seconds, captions burned in and kept inside the platform safe zone. Native vertical framing consistently outperforms cropped widescreen footage in feeds.
Ship episode one this week
You do not need a green-lit budget to enter this market — you need one hook, one locked character and one cliffhanger. Write the line, board the beats, and let the pipeline do the production work.
Popular guides