A subject in.
A finished video out.

Tell it what to talk about. It researches what's new, writes the script, has your presenter deliver it, weaves in scenes — and hands you one video, done.

🔎

Research

Fresh stories on your subject, gathered live

✍️

Script

A short script written in your voice

🧑‍💼

Presenter

Your avatar delivers it to camera

🎬

Scenes

Cinematic cutaways, generated to match

✨

One video

Assembled, ready to share anywhere

Keys stay in this browser tab only — sent per request, never stored on the server.

Commands

Everything the studio does is also scriptable from a terminal.

run news-avatar-video make a video: research → script → presenter → scenes → post run news-image-post make a quick text + image post --subject "…" what the content is about --period "past 24 hours" how fresh the research should be --sources 3 how many stories to draw from --duration 45 target video length, seconds --aspect-ratio 9:16 9:16 · 16:9 · 1:1 · 4:5 --language en output language --voice / --avatar presenter overrides --platforms tiktok,youtube where to publish --at 2026-07-09T14:00Z schedule instead of posting now pipelines list available workflows runs · show <id> see what ran, inspect every step accounts connected publishing accounts