Self-hosted · Open models · One GPU
Your own AI film studio.
One GPU. One click.
Images, video, character-consistent edits, camera angles and a full storyboard-to-timeline pipeline — running entirely on a GPU you rent by the hour. No subscriptions, no per-generation fees, no one else's servers touching your footage.
Runs on an RTX 4090 from ~$0.34/hr on Runpod community cloud — you pay Runpod directly, we take nothing.
How it works
-
1
Deploy
After clicking, on Runpod click a GPU card if none is selected (RTX 4090 cheapest that fits, RTX 5090 faster) and click Add volume (it shows the preset's size, which the template already sets) — that's where your models and films live.
-
2
Set a password
Under Set overrides, change
STUDIO_PASSWORDfromchange-meto your own. Left unchanged, a random one is printed in the pod logs. -
3
Open your studio
First boot downloads ~139 GB of models in the background — images are ready in about 3 minutes and everything in about 6–18 minutes, measured on a real RTX 4090.
Features
Studio
Free-form images, video, edits, camera angles and camera-motion presets — like Higgsfield, but on your own GPU.
Storyboard
Turn a script into scenes and shots with AI-assisted breakdown — characters, locations, blocking, action.
Location maps
Place cameras and actors on a top-down plan to get consistent, physically coherent camera angles every time.
Characters & LoRAs
Train or import LoRAs to keep characters, styles and locations consistent across an entire project.
Timeline & export
Arrange shots in scene order and export a normalized, concatenated MP4 — ready to share.
Your data, your pod
Everything lives on your pod's volume. Nothing is uploaded anywhere else to generate.
Models & licenses
Every model in the default stack is permissively licensed — you can use, modify and even sell what you make with it.
| Model | Role | License |
|---|---|---|
| Wan 2.2 (I2V A14B) | Video generation | Apache-2.0 |
| Z-Image Turbo | Fast text-to-image | Apache-2.0 |
| Qwen-Image-Edit-2511 | Multi-image editing & consistency | Apache-2.0 |
| Multi-angle LoRA | Camera angle control | Apache-2.0 |
| ComfyUI | Inference engine (unmodified, headless) | GPL-3.0 |
FAQ
Which GPU should I pick?
The RTX 4090 (24 GB) is the default and handles the full stack comfortably. A 5090 or L40S will generate faster if you don't mind paying more per hour.
Is my data private?
Yes. Everything — models, projects, generated media — lives on your pod's own persistent volume. Nothing is sent to a third-party API unless you optionally enable an LLM key for script breakdown.
What happens if I stop or terminate the pod?
Stopping the pod keeps the volume intact (you keep paying for storage, not compute). Terminating the pod deletes the volume — and everything on it — permanently.
What's the content policy?
The open models used here are not filtered beyond what upstream ships. The studio adds exactly one hard, non-optional block: sexual content involving minors is always rejected, with no setting to turn it off.