Automating clips and your own pipeline and your first script

The concept

Generation is offered through the studio and, on paid tiers, through an API for scripting; the useful automation is the pipeline around it — shot list, naming, assembly. The idea is simple: the same model you have been chatting with also answers a web request, so you can put it inside a script, a spreadsheet or a page. A key identifies you; a request sends the text; a response comes back as data.

The reason to try it once, even if you never build anything: it makes the chat version less mysterious. You see that the whole conversation is text in and text out, that your instructions are literally lines of a request, and that "the model" is one parameter among several.

A first call looks like this:

# The pipeline, and the parts no model does for you:
#
#  1. Write the shot list (subject / camera / light / duration).
#  2. Generate one reference still you are happy with. Keep it.
#  3. Generate each shot from that still, named shot-01.mp4 ...
#  4. Assemble in an editor: cut on movement, not on the beat of the music.
#  5. Add sound: room tone first, then a music bed, then narration last.
#
# Step 5 is why most generated video looks generated. Silent or
# music-only video reads as a demo; room tone makes it read as footage.
#
# ffmpeg, if you want the two-second version of "add room tone":
#   ffmpeg -i shots.mp4 -i room.wav -shortest -c:v copy out.mp4