OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Clip Guided Diffusion

Clip Guided Diffusion is a Python module and command-line tool for generating images from text with guided diffusion steered by OpenAI’s CLIP model. It supports weighted text prompts, image prompts, blending an existing image, and configurable generation settings; the repository is MIT-licensed.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Clip Guided Diffusion — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want a configurable, code-oriented workflow for CLIP-guided image generation, especially when you need prompt weighting or image-based guidance rather than a hosted, form-based image-generation service.

Who it is for

  • Developers seeking a Python-based image-generation tool.
  • Artists and experimenters who want to steer diffusion outputs with text or image prompts.
  • Researchers exploring prompt-driven image generation through a CLI or Python generator.

Documented uses

  • Generate an image from one or more text prompts.
  • Adjust the relative influence of prompt phrases with weights.
  • Guide generation with reference images or blend an existing image.
  • Save a generation as a GIF or MP4 and inspect intermediate outputs.
  • Integrate generation into a Python experiment using the provided generator.
Before you choose

Requirements & limits

  • Installation is documented through uv and a local repository clone.
  • The README says CUDA is selected automatically when available; CPU generation is supported but much slower.
  • Reducing timestep respacing can make generation quicker, with a documented trade-off in accuracy or alignment.
  • The CLI supports image sizes of 64, 128, 256 and 512.
  • Optional Weights & Biases logging requires an account.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the project and generate a text-prompt image

Set up the repository with uv and start a text-to-image generation run from the CLI.

Before you start

  • uv installed
  • A shell and Git
  • A CUDA GPU is recommended for substantially faster generation; CPU use is supported but documented as very slow

Steps

  1. Clone https://github.com/afiaka87/clip-guided-diffusion.git and change into the repository directory.

  2. Run the command below to install the project environment.

    Documented command
    uv sync

  3. Run the command below to start generation.

    Documented command
    uv run cgd -txt "Alien friend by Odilon Redo"

Expected outcome

The project starts generating an image; intermediate outputs go to `./outputs` and the current generation is saved as `current.png`.

Guide limitations

  • CPU generation takes a very long time compared with GPU use.
  • The README notes that bug fixes may be slow because development effort is being redirected to pyglide.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
4 October 2026
Original website title
afiaka87/clip-guided-diffusion · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

pyglide — The README says the author is redirecting development efforts to pyglide.

v-diffusion-pytorch — The README recommends looking at this project by @crowsonkb.