# CompVis/latent-diffusion

High-Resolution Image Synthesis with Latent Diffusion Models

Repository: https://github.com/CompVis/latent-diffusion
Canonical: https://ross.abutalabs.com/products/latent-diffusion
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2024-02-29T05:29:47+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1717, "days_push": 916, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 14133, forks 1733 (observed 2026-08-28T04:11:05.553142+00:00)

## What it is
The official research code and pretrained model zoo for Latent Diffusion Models (LDM), the paper behind Stable Diffusion, enabling high-resolution image synthesis in a compressed latent space. It provides Jupyter notebooks, scripts, and checkpoints for text-to-image, class-conditional, and retrieval-augmented diffusion generation.

## Use cases
- generate images from text prompts
- run pretrained latent diffusion models locally
- reproduce high-resolution image synthesis research
- experiment with retrieval-augmented diffusion models
- train or fine-tune diffusion models on custom datasets
- class-conditional ImageNet image generation

## When to choose
- you want the original reference implementation of latent diffusion models
- you need pretrained checkpoints from the LDM paper for research
- you want to study or extend the codebase behind Stable Diffusion

## When to avoid
- you want a production-ready or user-friendly Stable Diffusion tool - use diffusers or Automatic1111 instead
- you need actively maintained code with recent dependency support
- you lack a GPU or want CPU-only inference

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, stable-diffusion, data-science
- domain: deep-learning, machine-learning, artificial-intelligence, image-processing, computer-vision
- platform: python, cross-platform
- tags: diffusion-models, text-to-image, generative-models, research-code, pytorch, image-synthesis, gpu, linux

## Member repositories
- CompVis/latent-diffusion (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:05.553142+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:12:48.073446+00:00, confidence not recorded.
  - readme: https://github.com/CompVis/latent-diffusion (fetched 2026-08-28T04:11:05.553142+00:00, sha 4a9cb0191244)
- Data as of 2026-08-30T08:39:29.467469+00:00.