Script breakdown software for VFX, animation and live-action studios: every character, prop, set, vehicle and VFX shot, scene by scene — with what has to be built, how hard it is, and what the new draft just did to your number.
Free to start · no card · 30 days of everything from your first export
The breakdown is the product. But it is an unreleased script you are uploading, so the boring parts are handled properly and there is nothing here you have to configure to be safe.
Your files are encrypted on the way to us and while they sit with us.
Access, logging and change control designed around recognised industry standards.
Granular roles keep every draft in the right hands, and the log says whose.
Your content is never reused for anything else, and never used to train a model.
Read it, tag it, see the totals, hand it over. Four things, one screen each — no setup week, no pipeline integration, no training day.
Detect scenes, characters, props, locations, vehicles and VFX elements — an AI first pass that does the bulk of it.
Interactive boards, colour-coded elements and filters that answer "what has to be built" in one screen.
Work together securely across departments and locations, with roles, comments and an activity log.
Export production-ready reports in seconds — the workbook your estimate is built in and a PDF you can send a client.
Five steps, in this order — and the order is the point: reading a script to find every prop is work a machine is good at, and deciding what a prop costs is not.
PDF, Final Draft .fdx, Fountain or plain text. Scenes, slug lines, INT/EXT, day and night and the page numbers printed on the script are found for you, in English, German, Spanish, French or Italian.
The tool writes the prompt, you paste it into your own ChatGPT or Claude, and you paste the answer back. That single pass does the bulk of the breakdown, each element placed in the scenes and on the pages where the script actually says it. Nothing is accepted behind your back — every finding arrives as a suggestion you take or drop.
The read is worth doing because you are not hunting for nouns any more — you are adding what the page does not say: the asset that has to be doubled for the stunt, the coat that ages over three scenes, the vehicle that is really a hero build.
A second prompt goes back over what is now there: it fills in the descriptions that are still empty and says what looks missing or wrongly filed. You answer it the same way, one finding at a time.
One sheet per department in Excel, a CSV for whatever else the office runs on, and a PDF for the people who want a document, with your logo on it.
A VFX or animation house breaks a script down to price it — including the jobs it does not win. These are the parts that decide the number.
Tag "the coat" once and it is found on all nine pages it appears on, under all the names the writer gave it — quoted once, not three times, and not missed on page 84.
Rate an element on a scale your house defines per category — a prop is hero build or stock, a shot is 2D, CG or full CG. The summary totals what has to be built, per department.
Load the new draft next to the one you priced: pages, scenes, elements per department, assets to build and the ratings that changed. "Can we still hold the number" in one screen.
Live action and animation ship as presets and everything in them is yours to change. One tool for the CG series and the commercial you shoot next month.
What is tagged but not placed in the text, what is rated and what is not, what looks like a duplicate — the pass you would otherwise do at midnight before the quote goes out.
The application runs on a German server, and on Studio+ the database can be yours or sit in the region your legal department requires.
No print dialog, no "export to clipboard". A workbook your estimate is built in, and a document you can put in front of a client.
One sheet per department, page ranges collapsed the way a breakdown writes them, totals at the bottom. Add your rates in the next column and the estimate is done. CSV as well, for whatever your office runs on.
A real PDF, drawn page by page: a cover with your logo and the draft it was quoted from, the totals per department, a section for each of them, the scene list, and page numbers that say "3 of 12".
And never per export: a bid exported four times is a bid that was checked four times.
The 30 days start at your first export, not at sign-up — getting there is a day or two of real work, and a clock running through that is a clock spent on nothing. No card is asked for, so nothing can be charged when it ends. When it does, reading, tagging, both AI passes and the CSV export stay free forever.
Only if you want the AI passes, and then it is your own — you copy a prepared prompt into ChatGPT or Claude and paste the answer back. Nothing is sent anywhere by us, you are never billed for tokens, and every suggestion is accepted or rejected by you. Tagging by hand needs no AI at all.
In your organisation's own space, readable only by the people you invite. Scripts are not used to train anything. The application runs on a German server; on Studio+ the database can be yours, hosted by you or by us in your region.
PDF, Final Draft .fdx, Fountain and plain text. Scene headings, INT/EXT, time of day and page numbers are detected in English, German, Spanish, French and Italian.
Always. The CSV export is free forever, on every plan and after any trial has ended — a tool you cannot export from is a tool that owns your work. What a plan buys is the studio-format Excel workbook and the designed PDF.
They start the first time you export an Excel or a PDF, not when you sign up: getting from an uploaded script to a breakdown worth sending is a day or two of real work, and a trial that runs through that is spent on nothing. No card is asked for, so nothing can be charged when it ends.
Upload a draft and you have a breakdown worth sending the same day — every element, every mention, every number a quote is built on.
Upload your first script