Introducing Boring Studio

AI shot list generator

A shot list is normally paperwork.
This one films itself.

TheBoringApp generates a numbered shot list from one sentence: 2 to 6 shots, each with shot size, angle, lens, camera movement, light direction and the cast's fixed traits. Unlike a shot list you take to a shoot, every line here is directly filmable - the same text is the prompt that generates the clip. Writing and editing the list is free; only finished video is paid, from $0.80.

A shot list exists to answer one question before the day starts: what are we actually pointing the camera at, and in what order. Most AI tools that generate one hand you a table you then have to interpret. Here the list is the production - each row is the exact text that generates that shot.

What is on each line

Every shot carries a beat name so the structure reads at a glance - "Told he is fired", "Packing his box", "Leaving the office" - and underneath it, the shot itself in the language of camera and light: shot size and angle, lens and depth of field, the camera move, and the direction and quality of the light.

Each shot holds one continuous action. That is a hard constraint rather than a style preference: a shot that lists several actions in sequence makes the model morph between them instead of filming any of them, so "she turns her head toward the lens" is a shot and "she walks to the window, turns, and smiles" is three.

The list is ordered and the order is load-bearing. Shots film in sequence, each opening on the closing frame of the one before it, so reordering the list changes what continues from what.

Continuity is decided once, for the whole list

Before the shots are written, the cast and the look are settled: age, build, ethnicity, hair, wardrobe, distinguishing features, the location, the lighting and the grade. That reference is then restated verbatim inside every shot on the list.

This is the opposite of how a human shot list works, where you write the character's description once at the top and never again. The video model has no memory between shots, so anything named only once is reinvented on the next one.

A shot can be marked as a new scene, which breaks the chain on purpose: it stops continuing from the previous frame and gets an opening frame of its own. That is what you use when the story genuinely cuts - stepping outside, or jumping to later that day.

From list to footage

Approving the list puts every shot on a timeline as a draft. Nothing has been generated and nothing has been spent. You then film them one at a time, in order, seeing each finished 8-second clip before committing to the next.

Generating out of order is blocked rather than allowed-with-a-warning, because a shot that opens on the previous clip's last frame has nothing to open on if that clip does not exist yet - and you would have paid for the result before finding out.

A list can run to 10 clips on one timeline, which is 80 seconds of finished piece. The clips are stitched in order and downloadable with no watermark.

Questions

Is the shot list generator free?
Yes, and there is no credit card. Generating a shot list, editing any shot on it, reordering, adding and deleting shots, and starting over are all free and unlimited. You pay only when a shot is turned into video, bought outright from $0.80 per video with no subscription.
Can I use the shot list for a real shoot instead of generating it?
Yes, though it is written for a camera that has no memory rather than for a crew. Expect each shot to restate the character's wardrobe and hair, which reads as redundant on paper and is essential on screen. The beat names and the shot sizes, angles, lenses and lighting notes transfer straight to a human shoot.
How many shots does it generate?
Between 2 and 6, decided from the idea rather than from a number you choose. It splits on beats - one change of state per shot. You can then add or remove shots by hand, up to 10 on one timeline.
Can I change a shot after it is generated?
You can rewrite any shot's prompt and regenerate that shot. Because continuity is resolved at generation time rather than stored, regenerating an earlier shot means the ones after it pick up its new closing frame the next time they are generated - the list re-answers the question instead of holding a stale pointer.
Does it write camera moves that the model can actually do?
It is constrained to moves that survive generation: static holds, slow push-ins, gentle pans, handheld drift, a camera that flows from where the last shot left it. It will not write a crane move or a cut inside a shot, because a single clip cannot contain one.
What aspect ratios are supported?
9:16 vertical for TikTok, Reels and Shorts, or 16:9 landscape for YouTube. It is set once for the whole list, before any shot is generated, because every shot has to match. There is no square option - the video model cannot render it.

Free to try. Paid to render.

No card, no credits to decode. Write, rewrite and preview as much as you like - you only pay once you approve a still.