AnimateDiff

Motion added to an existing image model. The technique that made short animated clips possible on a home graphics card.

Self-hosted
Active
Allowed adult content
Open source

At a glance

Where it runs: Self-hostedYou run it on your own machine. Nothing leaves it unless you send it.
Adult content: AllowedAdult content is permitted within the published rules.
Trains on your content: NoYour content is not used to train models.
Status: ActiveMaintained and working today.
Account needed: No
Licence: Apache-2.0
Cost: Free and open source.
What it keeps: Local only.

What it takes to run

Graphics memory: 8 GB VRAM, 12 GB recommended
System memory: 16 GB, 32 GB recommended
Disk: about 20 GB
Runs on: Linux, Windows

Find what fits your machine on the tools page — pick your hardware from the first menu.

AnimateDiff inserts a motion module into a Stable Diffusion pipeline, so any checkpoint or LoRA you already have can produce a short animation rather than a still. It is a research project that became infrastructure: the ComfyUI and WebUI implementations are how most people meet it.

Clips are short — a couple of seconds — and coherence beyond that is still the hard part.

Worth knowing before you start

Clips are a few seconds Longer sequences drift and need stitching techniques.
Superseded for raw quality Dedicated video models beat it; it wins on running with what you already have.

What it does

Works with existing SD checkpoints and LoRAs
ComfyUI and WebUI integrations
Motion LoRAs

Models it runs

How it relates to other tools

requires ComfyUI — in practice, most people drive it from ComfyUI

Signs of life

Checked automatically. These are the only figures on this page a machine wrote, and they say when they were taken.

Last answered: Yes, 3 hours ago Its website responded when we asked.
Stars on GitHub: 12,246 1,095 forks.
Last commit: 777 days ago2024-07-31 01:14 UTC
Archived copy: Wayback Machine A snapshot, in case this one stops answering.