Earning From AI Short Video and Short Drama: The Complete Beginner's Playbook

For a hundred years, the route from a skill you have to money in your account ran through an employer. You proved the skill, someone hired you, and the company owned the equipment that turned your skill into output.

That route still exists — most of AIEH is about walking it well. But it is no longer the only one. A class of work now exists where the equipment is rented by the month, the crew is software, and the person with the idea keeps the output. This category covers that route: the services agentic companies sell to individuals, and how to actually operate them.

This is the playbook for TJP, an AI content platform. It is the clearest example of the pattern, because the gap between what it replaces and what it costs is the widest.

What TJP gives a complete beginner. Short drama is the hottest format on the internet right now, and TJP is the fastest way into it for someone with no skills at all — no camera, no crew, no editing training, and no prompt writing. Find a short video that is already working, copy it, change a few things: a few clicks and you have a video. Trim it, adjust it, and it is yours — your own short video, with your own character in it. From there it can start earning. And once the first one works, the next one costs you very little time.

Everything below is the detail behind that paragraph: which clicks, what they cost, what makes the result yours rather than a copy, and — stated honestly, because it is the part that gets oversold — how quickly the earning half actually arrives.

For the other services covered, and which suits which kind of work, see the directory of agentic services.

Data Notice: Vendor pricing, model names and feature availability in AI media tooling change on a scale of weeks. Tiers and prices below reflect the vendor’s published pricing at time of writing; treat the vendor’s own pricing page as authoritative. AIEH is not the vendor and does not resell this service.

Why an employment site covers this

AIEH’s argument is that hiring is moving from credentials to evidence — that what you can demonstrably do matters more than what a résumé claims. See skills-based hiring evidence for the research behind it.

Agentic services are the same shift, pushed one step further. If evidence beats credentials, then for some work you do not need the employer at all. You need the evidence and the production capacity — and the production capacity is now available for a subscription.

Three things follow, and they are the reason this category exists:

The cost floor of professional-looking media collapsed. Work that needed a crew, a camera package, a cast and an edit suite now needs one person, a subscription and an afternoon.
The skill that pays is operating the system, not owning the craft. You are not learning cinematography. You are learning to direct a machine that already knows cinematography — a different, faster, and much more teachable skill.
The same operation is hireable evidence. A channel you built and ran is a portfolio a recruiter can check in ninety seconds. Companies are already hiring for exactly this — see the operator and media roles on the AIEH job board.

What “a few clicks” actually means

The pitch for tools in this category is that a few clicks produce a video. That is true, and it is worth being precise about which few clicks, because the honest version is more useful than the marketing version.

TJP’s simple path is three steps: pick a look, add your talent, generate. That genuinely is a few clicks, and it genuinely does produce finished, publishable media. What it produces is a shot — one clip, a few seconds long, that looks like a production team made it.

Note what is not in those three steps: writing anything. You are not describing a scene from nothing. You are starting from a post whose settings already produced a good result, putting your own character into it, and changing what you want different. That is the whole beginner path, it takes minutes, and it removes the blank page — which is where nearly everyone who quits, quits. It has its own guide: copy, change, publish.

Two changes are what separate your video from a re-run of someone else’s: change the seed (otherwise you get the identical result back) and swap in your own character (otherwise the face belongs to a stranger and nothing compounds).

An episode is not one shot. A two-minute short-drama episode is somewhere between fifteen and forty shots, cut together, with sound. So the real sequence is:

Stage	What happens	Where
1. Script	A beat sheet and a shot list — the only writing you do	Anywhere
2. Cast	Lock a character so the same face appears in every shot	TJP avatars
3. Stills	Cheap test frames to settle framing, wardrobe and light	TJP
4. Shots	Animate the approved stills into clips	TJP video
5. Cut	Join clips, add sound and title cards	Any video editor
6. Publish	Post to the platforms that pay	TikTok, YouTube, Reels, short-drama apps
7. Read	Look at what performed, and make more of that	Platform analytics

Steps 1 and 5 are yours. Steps 2 through 4 are where the platform removes what used to be months of work. That is the whole compression, and it is enormous — but it is compression of production, not of judgment.

Each stage has its own guide in this category, linked below.

The feature that decides whether you have a business

Speed and cost get the attention. They are not the thing that matters most.

Character consistency is. An influencer’s actual asset is not the videos — it is recognition: a viewer knowing who this is before reading a word, and bringing to the post everything they felt about the last eleven. That asset has exactly one requirement, and it is that the person is the same person tomorrow.

Generic AI generation cannot do this, and no amount of prompt craft fixes it. Describe a person and you get a category; the model picks a new member of it every time. Run a channel that way and the audience meets eleven different people who share a hair colour, so post twelve inherits none of the trust post eleven earned.

The structural answer is an avatar — a character the models already know, with the identity locked while wardrobe, expression, pose and setting stay under your control. It is the reason a serial format is possible at all, and it is why the marginal cost of episode two is the shots rather than the casting.

It is also what a brand is buying. A brand paying for one video is buying production; a brand paying for access to a persona is buying an established relationship with an audience, which only exists because the identity persisted. The rates differ by an order of magnitude.

Two guides cover it: building a channel persona for why persistence is the asset and what quietly destroys it months in, and avatars: locking a character for the mechanics.

Where the money actually comes from

This is the part most tutorials get wrong, so it is worth stating plainly: an AI content platform is a production line, not a payout button. You do not earn by generating. You earn by publishing what you generated somewhere that pays, or by selling the capability to someone who needs it.

There are four realistic routes, and they pay on very different timelines:

Platform monetization. Ad revenue and creator funds on TikTok, YouTube, Instagram and the short-drama apps. Slowest to start — most programmes have follower and watch-time thresholds — but it compounds, and it is the one that scales without you.
Client and brand work. The fastest to first payment by a wide margin. A local business needs product video; you can now deliver it in a day at a cost that used to be a week and an invoice. You are selling delivery speed, not artistry.
In-platform creator earnings. Some services pay you when other users build on your published work. On TJP this pays in credits rather than cash, and it requires an approved creator application first.
Referral commission. Most of these platforms run a referral programme that pays real money on subscriptions you introduce. It is the only route here that pays without you producing anything, and it is also the one most likely to annoy an audience if you lead with it.

Where the money actually comes from works through each of these in detail, including thresholds, hold periods, and which are realistic in the first ninety days.

The honest expectation. The production step is genuinely fast — a first publishable clip in under an hour, a first episode in an afternoon. The earning step runs on the platform’s clock, not yours: audiences accumulate over weeks, and payout systems hold funds before releasing them. Anyone telling you the second step is as fast as the first is selling something.

What it costs to find out

Low enough that the question is answerable in a weekend rather than argued about.

TJP publishes three tiers: a free plan that generates on a pay-as-you-go basis, Pro at $50/month aimed at people posting weekly, and Max at $200/month for agencies and heavy users, which is also the tier that unlocks training your own avatar. Check the vendor’s pricing page before relying on any of those numbers.

The useful advice is the same on every platform in this category: use the free tier until you run out. Running out tells you which tier you need far more accurately than any comparison table, and it costs nothing to find out.

What an episode actually costs covers how generation is metered, why video costs many times what a still costs, and the drafting habits that cut a production bill by more than half.

The rights question, before you publish anything

Two rules cover most of it, and both are the vendor’s own stated position:

Likeness needs authorization. Generating a real person — a celebrity, a colleague, someone from a photograph you found — is permitted only where the likeness is owned, licensed or approved for use. Training a character on another person’s photographs needs that person’s agreement, because it builds something that can generate new pictures of them indefinitely.
Commercial use depends on your rights to the inputs. Your rights to the source media, the avatar, any brand or product in shot, and the final content all have to be clear before you sell or monetize it.

This is not boilerplate. It is the most common way a working AI video operation gets shut down after it has started earning. Rights, likeness and consent covers what is safe, what is not, and how to keep a record that shows the difference.

Start here

If you have never generated anything, the guides in this category are ordered to be read in sequence:

Copy, change, publish — the fastest path for a complete beginner. No prompt writing at all.
TJP quickstart: from nothing to your first clip — the from-scratch path, for when what you need does not exist yet.
Writing a short-drama script an AI tool can shoot — the beat sheet and shot list that make the rest mechanical.
Avatars: locking a character — the mechanics of the feature everything else depends on.
Building a channel persona — why identity persistence is the asset, and what breaks it months in.
Prompting for video: motion, camera and light — why video prompts fail, and what to write instead.
What an episode actually costs — the metering, and how to spend a fraction of it.
From clips to a finished episode — continuity, cutting and sound.
Covers and thumbnails — the frame that decides whether anything else mattered.
Producing a season, not a clip — batch discipline once one episode works.
Where the money actually comes from — the four routes, with real timelines.
Rights, likeness and consent — before you publish, not after.
When a generation fails or looks wrong — the failures everyone hits in week one.
Turning an AI video operation into hireable evidence — because the channel is also a portfolio.

If you would rather be hired for this

Running a channel and being paid a salary to run one are both real options, and they are not in conflict — the first is the qualification for the second.

Companies are hiring operators, prompt engineers, brand directors and social managers for exactly this work right now, and several are listed on the AIEH job board. If that is the direction you want, build the operation first and then make it legible: turning an AI video operation into hireable evidence covers how to convert a channel into something a recruiter can verify, and what is the Skills Passport covers the credential that carries it between employers.

Last reviewed: 2026-08-23 · Editorial policy · Report an error