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

Mflux

Mflux is an open-source Python library that implements generative image and video models in Apple MLX, with documented command-line and Python workflows for image generation.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Mflux — 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 Mflux when you want a Python library and documented CLI for running generative image models with Apple MLX, particularly in a local Mac workflow.

Who it is for

  • Developers building image-generation applications or workflows around MLX implementations.
  • Mac users who want to run supported generative image models locally.
  • Python developers who want to generate images through a library rather than only a command-line workflow.

Documented uses

  • Install Mflux and generate an image from a text prompt with a model-specific CLI command.
  • Call a model from Python, configure generation parameters and save the resulting image.
  • Choose among listed model families based on documented capabilities such as editing, upscaling or text rendering.
Before you choose

Requirements & limits

The repository is public and specifies the MIT license. The README documents installation with uv.

  • Mflux implements models in MLX and uses Hugging Face Transformers tokenizers.
  • The first run downloads the selected model, which may take time.
  • Model families have different strengths; the repository directs users to model-specific READMEs for full usage details.
  • The README gives Mac as the local use case and also shows a separate uv installation command for DGX/NVIDIA, without specifying support details for that route.

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 Mflux and generate a first image

Install Mflux with uv and use the documented Z-Image Turbo command to generate an image.

Before you start

  • Install uv.
  • A Mac is the documented local use case.
  • The selected model downloads automatically on its first run and may take time.

Steps

  1. Install or upgrade Mflux with the command below.

    Documented command
    uv tool install --upgrade mflux

  2. Run the command below to see available Mflux CLI commands.

    Documented command
    uv tool list

  3. Run mflux-generate-z-image-turbo with a prompt and the documented width, height, seed, steps and quantization options.

Expected outcome

Mflux generates an image from the supplied prompt using the selected model and settings.

Guide limitations

  • The example covers Z-Image Turbo; consult the model-specific README for other models and their usage details.
  • The first run downloads the model and may take some time.
Documentation 1 ↗
Source-based practical guide

Generate and save an image from Python

Use the documented Python API example to create and save an image with Z-Image Turbo.

Before you start

  • Install uv.
  • Use Python 3.10 or later, as specified by the example's inline script metadata.
  • A Mac is the documented local use case.

Steps

  1. Create a Python script with inline uv metadata listing mflux as a dependency.

  2. Import ZImageTurbo from mflux.models.z_image and construct the model with the example's quantization setting.

  3. Call generate_image with the documented prompt, seed, inference steps and dimensions.

  4. Save the returned image with image.save.

Expected outcome

The script generates an image and saves it to the specified filename.

Guide limitations

  • The example covers Z-Image Turbo; other models may use different APIs or settings.
  • The first model download may take some time.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
mflux-community/mflux · Official homepage title · retained in its source language
How we review sourcesSuggest a correction