Short-form video, rendered on your own GPU

Two ways to use it

The same pipeline is available two ways, and they trade speed for effort in opposite directions.

Install

$ pip install video-gen-local
$ vgl install       # GPU environment + models, ~1 hour
$ vgl serve         # studio on localhost:7860

That is the whole install — no clone, no checkout. The setup scripts ship inside the package, so vgl install runs from wherever pip put it.

Why two steps and not one

pip install gives you the CLI, the studio, the renderers, text-to-speech and captions. It cannot give you the avatar model, for two reasons that are not going away:

• It needs Python 3.10 where the rest of this needs 3.12, so it lives in its own environment.
• Its flash-attn wheel has to match the C++ ABI of the torch that was just installed. A dependency list cannot say "whichever wheel matches the torch you are about to have".

And the weights are about 110 GB of model downloads, which are not Python packages at all — pip was never going to fetch those. vgl install does both jobs, in the order that matters (torch before the requirements that would overwrite it, numpy re-pinned last).

Prefer to work from source? git clone then pip install -e . gives you the same three commands.

What you need

Runs the same on your own machine, on Nebius, or on AWS. The only real difference between providers is whether stopping the machine wipes its disk — on a VM cloud it does not, on RunPod or Vast it does, and vgl install simply re-provisions.

Keys

Every account is your own. There are no shared credentials in the source, and the studio refuses a job whose template needs a key you have not set rather than failing partway through a render.

$ cp .env.example .env # required keys are at the top, uncommented

Make one

Open the studio, pick talking-head, drop in a front-facing photo, type a ticker and a sentence of notes, press Generate. About fourteen minutes at 480p.

General videos: every template can render with no identity bar at all — no handle, channel, tier band or ticker chip in the frame. Tick General video in the studio, or pass is_general: true on the API. Publishing to a Variant channel with asChannel implies it.

Templates

The layouts. They differ in what they put on screen and what they need from you — not in quality.

catalyst A ~10s alert. Presenter on a solid field under a looping graphic. ticker · notes or events feed

not yet rendered
on this machine

catalyst-split The alert, split in half: an establishing shot generated for the event loops on top, the presenter speaks on a soft blue gradient below, a black channel bar riding the seam. The script pass writes the shot's visual direction in the same breath as the alert. ticker · events feed · FAL_KEY

showdown A three-hander debate. Narrator, bull and bear, each with their own face and voice. insights feed · three faces

showdown-no-lipsync A fight card. Generated art per round, claims spoken over it, no lip-sync. insights feed · two faces · FAL_KEY

not yet rendered
on this machine

recent-event A filing, explained. Headline card, presenter over a chart on a forest-green field, animated points, and a quote copied verbatim from the SEC filing with the executive who said it. Writes itself; a blank ticker covers the newest confident catalyst in the feed. nothing required · optional ticker · HEDGEFUND_API_KEY

not yet rendered
on this machine

earnings-call The latest earnings call, on a violet field. The narration and on-screen points are written from management's prepared remarks in one pass, and the quote is a sentence lifted from the transcript with its named speaker — never composed. A blank ticker covers the newest call on a US listing. nothing required · optional ticker · FMP_API_KEY

not yet rendered
on this machine

social-clip Presenter over a live stock chart, with an optional X post card above it — or, with no ticker at all, the post alone carries the clip and the narration reacts to it. ticker or X post · one face

not yet rendered
on this machine

chart-head The talking head over the ticker's live chart, on a slate field — navy gradient page, matching dark chart card, white identity bar. Every spoken word is yours. ticker · one face · notes or script

not yet rendered
on this machine

talking-head Presenter only, over a flat background. The simplest thing that works. one face · notes

not yet rendered
on this machine

talking-head-split The split frame with your words: a generated establishing shot loops on top, the presenter speaks below on a colour you pick (blue, green, red, black, purple, orange — always the same pale-into-deep gradient), the channel band on the seam. No opener, no feeds. one face · notes or script · optional ticker · FAL_KEY

not yet rendered
on this machine

trade-video A move announced on location: the face is re-staged into a scene — beach, limo or penthouse — before it is animated, so the whole frame is the shot. The alert line is written from your trade facts, your notes, or spoken verbatim from your script. one face · script, trade facts or notes · FAL_KEY

These are frames from real renders on this machine, not mockups. The two without one have simply never been run here — the studio shows every template regardless.

How it works

Variant integration

Access note: publishing finished videos into Variant uses Variant’s pre-made publish endpoint, which is a restricted capability — your Variant account needs publish access granted in addition to an API key. Request it from Variant; rendering works either way, only the publish step needs the grant.

If you publish to Variant, the studio can render as one of your avatars and post the finished video back. Your token is the identity — paste it once and everything else is read from it.

If something breaks

vgl doctor --deep checks every stage and tells you which one is unhappy.