How the docs are read

For each model, the maker's own prompt guide, API reference or product page was read on one day. Every statement about camera moves, framing, cuts, frames, negative prompts or motion was copied with a short quote and given one reading from a closed list. As of 2026-09-25.

1Which models are on the register

A model is listed when its maker publishes documentation that says something about directing the camera, framing or frames. Models whose only public page is a showcase were left off; so were hosting services that carry other makers' models without documenting camera language of their own.

2Camera words

The 28 camera words were searched for in the pages read. A word counts for a model when the docs define it, list it as a command or preset, or use it in their own example prompts. A word that appears only in unrelated text, a head tilting or a pedestal under a product, was not counted.

3Unreadable pages

When a maker's reference asked for a sign-in or showed no text, the point is recorded as unreadable, not as missing. That is the case for one model's API reference on the day read.

4Readings in use

Inclusion rule. Readings actually given on the register. Order. Fields in register order.

The full register is on the data page.

5Camera language in a pipeline

Where “How docs are read” has to be written by hand, SceneMixer's pipeline writes the shots; its guides explain:

Paste a scene or a short script and SceneMixer gives you a storyboard — numbered shots with framing and seconds — and then a moving version of each segment, which you can join into a rough cut of the scene.

SceneMixer, Using SceneMixer for Previs and Animatics Before a Live-Action Shoot (read 2026-10-13)

SceneMixer writes dialogue into the script analysis, assigns each line to a shot with a speaker and a short delivery note, and the video model performs it — picture and audio generated together, so the mouth forms that sentence in the language you chose.

SceneMixer, AI Dialogue Generator (read 2026-10-13)