# CueTheScene: full context
> Long-form YouTube videos from a topic prompt: script, voiceover, footage, captions, and a thumbnail, ready to publish.
CueTheScene is an AI video production tool for long-form, faceless YouTube creators. You give it a topic and a format, walk away, and come back to a publish-ready video: script, voiceover, footage, captions and a thumbnail, with a publish kit ready for YouTube Studio (video file, title, description, tags, captions, thumbnail). It is built for long-form, faceless channels in niches like history, true-crime, science, explainer and commentary, and ships in two formats, documentary and listicle, not short clips.
Built and run by MARKETEZEAI LTD, a UK company (Companies House: https://find-and-update.company-information.service.gov.uk/company/16787865). The founder previously produced the YouTube channel "WW2: As It Happened" (https://www.youtube.com/@WW2AsItHappened) with a different tool. That is operator pedigree, not output from this product. In test now; launching July 2026. Canonical site: https://cuethescene.com.
## Pricing
Four plans. The whole pipeline (script, voiceover, footage, captions, thumbnail, and a publish kit for YouTube Studio) is on every plan; what changes is monthly video volume, concurrency, team seats and storage. The first 100 founding members get 30% off for life. New subscriptions and plan changes use monthly billing. Yearly billing is not available yet.
- **Starter** — £59/month. Built for the creator shipping about one video a week.
- **Creator** — £149/month. Built for three uploads a week, one channel, no editor.
- **Pro** — £449/month. Built for a daily upload, or two channels run solo.
- **Scale** — £1199/month. Built for an agency or a multi-channel operation with a team.
### Billing questions
### Can I pay yearly?
Not at the moment. New plans and plan changes use monthly billing. If you already have a yearly plan, it keeps running, and you can switch to monthly from account settings.
### Can I cancel any time?
Yes. There is no contract and no lock-in. Cancel from your account settings and your plan keeps running until the end of the period you have already paid for, then stops. Nothing renews after that.
### What happens if I change plan partway through a month?
Upgrades take effect immediately and the new credit allowance is available straight away, with the change reflected on your next invoice. Downgrades take effect at the start of your next billing period so you keep what you have already paid for.
### Do you do refunds?
We do not do cash refunds. If something fails through a fault on our side, say a render that breaks because of a bug in our pipeline, a sustained outage, or a billing mistake we made, we put a service credit on your account to make it right. We do not refund a video you did not like, views or monetisation that did not happen, misuse of the tool, or an outage at a third party such as the YouTube API. Your statutory rights under the UK Consumer Rights Act 2015 are untouched and always apply.
### I've already generated videos and want to stop. Where does that leave me?
Generating a video uses up the cancel-right for that video only, not your whole plan. Within 14 days of buying you can still cancel and get back the part of the fee that covers the credits you have not yet used. Once the 14 days are up, you can cancel any time for future periods but the time you have already paid for is not refunded. Your statutory rights under the UK Consumer Rights Act 2015 always apply.
### Are prices in pounds? What about tax?
Every price on this page is in GBP. Tax is added at checkout based on where you are. Our payment provider handles the tax side, so you see the final figure before you pay.
### How do I get a receipt or an invoice?
Your receipts and invoices come from Polar, our payment provider, and sit in your billing portal. Open Settings, then Billing, then Manage billing and receipts to download any of them. If you need a company name, VAT number, or billing address on the invoice for your records, you can add those in the portal.
### What is the founding-member price?
The first 100 customers get 30% off their monthly plan price for life. It does not expire and it does not step back up. Once the 100 spots are gone, that is it.
### How do the video counts work?
Each plan is a monthly pot of credits. A typical 12-minute video is what the "videos a month" figures are based on. A shorter video costs less and a longer one costs more, so the real number flexes with what you make. If you go over, top-up packs cover the overflow without changing your plan.
## Features
### Real-time generation status
See every step of the render as it happens — script, voice, footage, captions, thumbnail.
You should never wonder whether the render is stuck or just thinking. The status panel shows every phase of the pipeline as it runs, with a clear next-step and an estimate of how long is left.
**What it is.** A live status rail down the side of every job: outline → script → voiceover → footage → captions → thumbnail → publish. Each step shows its current state, its duration, and any soft-warning the pipeline flagged. If something needs your eyes, it goes amber rather than rolling on quietly.
**Why it matters.** AI video tools usually go quiet for ten minutes and then either succeed or fail. That makes it impossible to plan around. You either babysit the tab or risk re-running a 60-minute job because a single asset 404'd. We surface the work so you can walk away with confidence.
**Who it is for.** Faceless creators producing two-plus videos a week — the people for whom "is it done yet?" is a real productivity question, not a curiosity.
**In practice.** A render at the 45% mark might read: "Script reviewed. Footage curating: 18 of 24 scenes. Voiceover starting in ~90 seconds." You always know where you stand.
### Per-scene regeneration
Fix one scene's footage or narration without touching the rest.
The third scene reads wrong. Fix just that scene. The other 23 keep their voiceover, footage, captions and timing. Then render the video once, with your edits.
**What it is.** Every scene is its own block in the editor. Pick fresh footage for scene 3, search for a better clip, or use your own image. Edit its narration in the captions and re-voice only that scene. The change shows in the preview, and you can undo it or try again until it lands.
**Why it matters.** The "regenerate everything" model is how most AI video tools work, and it's why creators end up with one almost-right video and no good options. Per-scene regen turns a "this is broken" moment into a 30-second fix.
**Who it is for.** Anyone shipping long-form video where a single off scene is the difference between a publish and a rewrite.
**In practice.** Common use: the AI picked a generic stock clip for "the eve of D-Day". You open that scene in the Media panel, search for "1944 Normandy" and swap in an archive clip. The rest of the video is untouched.
### Your YouTube publish kit
Every finished render comes with a kit: the video file, title, description, tags, captions and thumbnail.
When a render lands, everything you need to post it is right there. No retyping the title into YouTube, no hunting around for the description.
**What it is.** Every finished video comes with a publish kit. Download the MP4, thumbnail and caption file, and copy the title, description and tags into YouTube Studio in a couple of clicks. The description already has your footage and music credits and the AI-made disclosure baked in, so a straight copy-paste keeps you above board.
**Why it matters.** Writing a title, description and tags from scratch is a fifteen-minute job every single video. The kit hands you all of it, ready to paste, and there is no channel access to grant, so you stay in control of what goes live and when.
**Who it is for.** Anyone who already publishes through YouTube Studio and wants the metadata written for them, without handing a tool the keys to their channel.
**In practice.** Render finishes, you open the kit, copy the description (credits and disclosure already in it), grab the MP4, thumbnail and captions, and upload in YouTube Studio. Two minutes, not twenty.
### Multi-source footage
Pexels, Pixabay, NARA, Library of Congress, Internet Archive — routed by niche.
A history channel should pull from period-correct archives. A finance channel should pull from clean, modern stock. We route the footage hunt by niche so the clips you see fit the channel you ship to.
**What it is.** A niche-aware orchestrator queries the right sources first: NARA, the Library of Congress, the Internet Archive, the Smithsonian, and Wikimedia Commons for historical pieces; Pexels and Pixabay for modern documentary and tech. Each query is vision-checked before the clip lands in your timeline so a "1940s tank" prompt doesn't come back with a museum diorama.
**Why it matters.** Generic stock libraries are why AI documentary footage tends to look generic. Niche routing is what turns a finished render from "good enough" into "channel-fit".
**Who it is for.** Anyone publishing in a niche where the look of the footage is half the perceived quality — history, science, biography, finance, tech.
**In practice.** A WW2 documentary scene about Anzio (1944) returns three candidates: a NARA archive clip, an Internet Archive newsreel cut, and a British Pathé still. You pick the lead.
## Frequently asked questions
### What is CueTheScene?
You give it a topic and a format, you walk away, and you come back to a finished long-form YouTube video: script, voiceover, footage, captions and a thumbnail, ready to publish. It is built for faceless long-form channels, not fifteen-second clips.
### How long does a video take to generate?
It depends on the length and on how much footage the job has to search. You do not sit and watch it. You start the job, close the tab, and we email you when the video is ready to publish or tweak.
### What niches does it work best for?
It is strongest where archival and stock footage carry the story: history, documentary, true-crime, explainers, list videos and commentary. It ships in two formats today, documentary and listicle, so the honest answer is to try your topic and judge the output. The first video tells you more than any promise here would.
### Can I edit the result, or is it a black box?
You can edit all of it. There is a full editor: rewrite any line, regenerate a single scene, swap a clip, change the music, restyle the captions, then re-render. The point is that you do not have to. It comes out publish-ready, and the editor is there for when you want it, not because you need it.
### Can I use my own script?
You are in full control of the script. CueTheScene writes a first draft from your prompt, then you edit it line by line in the editor before a single second is rendered. If you would rather narrate it yourself, you can upload your own voiceover instead of using a generated voice.
### What about custom voices?
Pick from a large library of natural voices, or use your own cloned voice so every video sounds like the same narrator. You can also upload a voiceover you recorded yourself and build the video around that.
### What about copyright on the stock footage?
Every clip and track we pull is either public domain or licensed for commercial use, including monetised YouTube. Footage comes from stock and archival libraries that allow this. Music is a self-hosted library released under CC-BY, and the attribution is added for you. You should not get a copyright claim because of media we sourced. Some archival footage is Share-Alike (CC BY-SA), which means a video using it is itself shareable under the same licence, so others can remix it. We tell you per video when that applies, and you can prefer clean-ownership footage in your settings if you want exclusive ownership.
### Is my data private?
Yes. Your prompts, scripts and videos are yours. We do not train models on your content. You can export everything or delete your account and all of it from your settings, and we action that properly, not just hide a row. The privacy policy spells out exactly what we hold and why.
### What is included in each plan?
The whole pipeline is in every plan: script, voiceover, footage, captions, thumbnail and a publish kit for YouTube Studio. What changes between plans is how many videos you ship a month, how many run at once, and upload storage. The pricing page lays the four plans out side by side.
### Can I add team seats?
Not yet. Every plan today is a single-operator account. Team workspaces and seats are coming soon, aimed at agencies and multi-channel operations. For now, if you are running one channel on your own, you have everything you need.
### Can I cancel any time?
Yes. There is no contract and no lock-in. Cancel from your account settings and your plan keeps running until the end of the period you have already paid for, then stops. Nothing renews after that.
### Do you do refunds?
We do not do cash refunds. If something fails through a fault on our side, say a render that breaks because of a bug in our pipeline, a sustained outage, or a billing mistake we made, we put a service credit on your account to make it right. We do not refund a video you did not like, views or monetisation that did not happen, misuse of the tool, or an outage at a third party such as the YouTube API. Your statutory rights under the UK Consumer Rights Act 2015 are untouched and always apply.
### Can I publish to multiple YouTube channels?
Yes. There is no channel to connect. You get a publish kit for each finished video and upload it yourself in YouTube Studio, so you can post to as many channels as you like. Running several channels is exactly what the higher plans are sized for.
### Do you support Shorts?
Not at launch. CueTheScene is long-form first, that is the harder problem and where most faceless channels make their money, so the two launch formats are documentary and listicle. Vertical Shorts are on the roadmap, not in the launch product, so do not pick a plan expecting them yet.
### Do you have an API?
Not at launch. It is on the roadmap but it is not in the launch product, so do not pick a plan expecting one yet. If an API is the thing that matters to you, email us and tell us what you would build with it. That genuinely shapes what we do next.
### Where do I get support?
Email hello@cuethescene.com and a real person, usually the person who built this, gets back to you, normally within one working day (UK time). There is no outsourced support desk and no bot wall to argue with first. If something is urgent, a render stuck or you cannot get into your account, say so in the subject line and it jumps the queue.
## Blog
### A solo creator's guide to staying monetised on YouTube in 2026
Published 2026-05-15. What the 2026 monetisation rules actually ask of faceless channels, and how to keep an AI-assisted workflow firmly on the right side of them.
The question I get most often is some version of "will an AI-assisted video
get me demonetised". It is the right thing to worry about and the answer is
not a slogan. Here is how I think about it as someone running a channel, not
selling you peace of mind.
## What YouTube actually cares about
Strip the policy language back and it comes down to one idea: is this
genuinely your content, made to be watched, or is it mass-produced filler made
to game the algorithm. Channels get into trouble for the second one. The
distinction YouTube keeps drawing is between content that is authentic and
has a point of view, and content that is repetitious, low-effort and
inauthentic.
That line does not run along "was AI involved". It runs along "is this
actually a video someone made on purpose". Plenty of fully hand-edited
channels are on the wrong side of it, churning out the same template daily.
Plenty of AI-assisted channels are firmly on the right side of it, because the
creator has a niche, a voice and an editorial hand on the output.
## Where AI-assisted workflows go wrong
Not because they used a tool. Because of how they used it:
- **Volume with no editorial hand.** Forty videos a week, none of them
watched by the person who published them. That is the pattern that gets
flagged, and it is a choice, not a consequence of the tool.
- **No niche.** A channel that does whatever topic is trending today reads as
a content farm, hand-made or not.
- **No point of view.** If every video could have been made by anyone about
anything, there is nothing authentic for the channel to stand on.
None of those are AI problems. They are channel-strategy problems that AI
makes faster to commit at scale. The tool is an amplifier, not an excuse.
## How to stay on the right side of it
Treat the pipeline as the assembly, not the editorial. That means:
1. **Own a niche.** Depth in one area reads as authentic. The
[niches that genuinely work](/blog/the-15-niches-where-ai-documentary-works-in-2026)
are a good place to start.
2. **Watch your own videos before they go out.** If you would not sit through
it, do not publish it. Per-scene editing exists precisely so you can fix
the bits that fail that test instead of shipping them.
3. **Have a voice.** Your script, your angle, your through-line. The draft is
a starting point you edit, not a thing you publish unread.
4. **Pace your uploads like a creator, not a farm.** Consistent and
considered beats forty-a-day every time, for the algorithm and for the
reviewer.
## The honest summary
An AI-assisted documentary is not a monetisation risk because it was
AI-assisted. It is a risk if it is inauthentic, repetitious, off-niche and
unwatched, and it is exactly as much of a risk hand-edited. Use the time the
pipeline gives you back to be more editorial, not less, and you are building
the kind of channel YouTube wants more of, not less.
That is the whole approach the tool is built around. [What it does](/#features)
covers the editorial control, and [pricing](/pricing) is sized so you can
keep a sane, considered upload schedule rather than a farm one.
Cheers,
Carl
---
### Why we built per-scene regeneration
Published 2026-05-11. Regenerating a whole video to fix one weak scene is the wrong unit of control. This is why the editor works scene by scene.
Early on, the regenerate button worked the way you would expect: you did not
like the video, you pressed it, you got a different video. It tested well in a
demo and it was the wrong design. Here is why we tore it out.
## The unit of control was wrong
A ten-minute documentary is maybe forty scenes. When scene 31 has the wrong
footage, the other thirty-nine are fine. Regenerating the whole thing to fix
one of them throws away thirty-nine good scenes to re-roll one bad one, and
hands you a completely different video where the one problem might now be in
scene 12 instead. You were not editing. You were pulling a slot machine and
hoping.
That is the actual complaint behind "the AI video tool gave me something I
could not use". It is not that the output was bad. It is that there was no way
to fix the one thing wrong with it without risking everything that was right.
## Scenes are the natural unit
So the unit of control became the scene. Scene 31's footage is wrong: you
regenerate scene 31. The narration is good but the clip is generic: you swap
the clip and keep the narration. The line is clunky: you rewrite the line and
re-record just that. Forty scenes, forty independent dials, and the
thirty-nine you were happy with never move.
This is the difference between a tool that produces videos and a tool you can
actually finish a video with. The first one is a demo. The second one is a
workflow.
## It changes what "AI-assisted" means
When the only control is regenerate-everything, you are not really involved in
the video. You are a judge giving it a thumbs up or down. When the control is
per-scene, the video is yours: the structure, the footage choices, the
phrasing, the pacing, all editable in place. The pipeline did the assembly so
you could spend your attention on the decisions that actually change the
result.
It also changes the economics. Re-rolling a whole render to fix one scene is
expensive and slow. Regenerating one scene is neither. The control got finer
and cheaper at the same time, which is usually a sign the design is right.
## What this looks like in practice
You get a finished video. You watch it once. Three scenes are not pulling
their weight. You regenerate those three, swap one clip you would have chosen
differently, tighten two lines, and re-render. Ten minutes, not a re-roll,
not a rebuild, and the thirty-seven scenes you liked are exactly as they were.
That is the whole argument. The editor is built around it end to end, and
[what it does](/#features) walks through the rest of it. If you want to see
what it costs to work this way, that is on the [pricing](/pricing) page.
Cheers,
Carl
---
### The 15 niches where AI documentary works in 2026
Published 2026-05-07. Not every topic carries on archival and stock footage. Here are the niches where it genuinely works, and the ones where it does not yet.
An AI-assisted documentary lives or dies on one thing: can the story be told
in archival and stock footage. Where it can, the result is genuinely good.
Where it cannot, no tool saves it. So here is the honest split, not the
everything-works pitch.
## What makes a niche work
Three things. The story is mostly narration over imagery, not a person talking
to camera. The imagery exists in public-domain or stock archives. And the
topic rewards depth, so a ten or twenty minute video is the right length
rather than a stretch. When those three line up, the pipeline does its best
work.
## The 15 that work
1. **History**: the strongest of all. Real archival film, not a re-enactment.
2. **War history**: footage archives are deep and the narration carries it.
3. **True crime**: structure and pacing matter more than original footage.
4. **Space and astronomy**: agency imagery is public domain and stunning.
5. **Aviation and maritime disasters**: investigation-shaped, archive-rich.
6. **Biographies**: a life is a narration with a clear arc.
7. **Ancient history and archaeology**: slow, deep, image-led.
8. **Science explainers**: concept plus diagram plus narration.
9. **Economics and finance explainers**: abstract ideas made concrete.
10. **Geography and geopolitics**: maps and footage do the explaining.
11. **Engineering and megaprojects**: process stories with strong visuals.
12. **Nature and wildlife**: stock libraries are deep here.
13. **Mysteries and unsolved cases**: question-led, retention-friendly.
14. **Technology history**: how we got here, told over archive.
15. **Debunking and myth-busting**: argument-led, evidence on screen.
Notice the pattern: every one of these is a story you narrate over footage you
did not personally shoot. That is exactly the shape the pipeline is built for.
## Where it does not work yet
Be honest with yourself about these, because picking the wrong niche is the
fastest way to be disappointed.
- **Anything that needs your face.** Vlogs, talking-head commentary, reaction.
If the value is you on camera, this is the wrong tool.
- **Fast news reaction.** If the video is worthless an hour later, a
twenty-minute considered pipeline is the wrong fit.
- **Highly current commentary.** Topics that move daily move faster than a
deep edit deserves.
- **Anything needing footage that does not exist in an archive.** A niche
about a brand-new product has no public-domain library behind it.
## How to choose
If your topic is in the first list, you are in good shape, pick the niche and
go. If it is in the second, the tool is not going to fix that, and a tool that
told you otherwise would be lying to you.
The full set of formats and niches is on the [features](/#features) page, and
[pricing](/pricing) shows how many videos a month each plan covers so you can
match it to your upload schedule.
Cheers,
Carl
---
### How CueTheScene writes a 10-minute documentary
Published 2026-05-02. A walk through every stage between a one-line topic and a finished, publish-ready documentary, and where you can step in if you want to.
Most of the work in a long-form documentary is not the writing. It is the
twenty other jobs around the writing: finding footage that matches the
narration, timing the voiceover, cutting it together, captioning it, making a
thumbnail that earns the click. That is the part that eats your evening. Here
is what actually happens between you typing a topic and getting a finished
video back.
## You give it a topic and a format
You write one line, the way you would describe the video to a mate. You pick a
format, say Documentary, and a niche. That is the whole brief. Everything below
runs without you.
## It writes the script, then argues with itself
A first draft gets written from your prompt and the sources gathered for it.
Then a separate critic reads that draft back against the sources: does each
claim have a source behind it, does the story hang together, could any line
harm a real person. Weak scenes get rewritten from the sources, and the critic
reads them again. A script that still fails stops there, before any voiceover
or render.
## It breaks the script into scenes
The script is split into scenes, each with its own job: this one sets up the
question, this one pays it off. Every scene gets a length and a description of
what should be on screen while the narration runs. This is the part that makes
the footage match the words instead of being generic B-roll.
## It finds footage that fits
Each scene goes looking for footage across stock and public-domain archives:
national archives, the Library of Congress, the Internet Archive, Wikimedia,
and stock libraries for the modern shots. For a history piece that means real
archival film, not a stock clip of someone pointing at a map. Everything it
pulls is cleared for commercial use, including monetised YouTube, so you are
not inheriting a copyright claim.
Music is the same story: a self-hosted library released under CC-BY, with the
attribution written for you. Nothing in the finished video is a licensing
problem waiting to happen.
## It records the narration
The script becomes a voiceover in a natural voice you picked, or your own
cloned voice if you set one up. The narration is checked for the usual
text-to-speech failures, the mangled word, the wrong emphasis, the half-second
of silence in the wrong place, so you are not the one catching them on
playback.
## It captions, renders and makes a thumbnail
Captions are generated and burned in. The footage, narration, music and
captions are assembled into the finished render. A thumbnail is generated to
go with it. Then a final pass reviews the rendered video as a whole, the way a
viewer would, and flags anything that came out wrong.
When it is done, you get an email and a publish-ready video.
## Where you step in, if you want to
Here is the important bit: none of the above needs you. But all of it is
yours to change. Every scene can be regenerated on its own. You can rewrite a
line, swap a clip, change the music, restyle the captions, then re-render.
Then you open the publish kit and upload it in YouTube Studio.
The point of the tool is not to take the editing away from you. It is to make
the finished video the starting point instead of the destination, so the
control you spend it on is the control that actually changes the video, not
the hours of assembly that never did.
If that is the workflow you want, the [pricing](/pricing) page lays out what
it costs, and [what it does](/#features) goes through the editor in more
detail.
Cheers,
Carl