Skip to content

Flex.2-preview

An open text-to-image model with inpainting and control

Visit official site

What Flex.2-preview is

The model is described as an open-source text-to-image diffusion model with built-in inpainting and general control, supported by a community project. It is stated to serve many creative projects and to improve through user feedback.

What you can do with it

  • Inpaint and edit existing images with the built-in feature
  • Use pose, line and depth inputs for general control
  • Fine-tune the model for custom needs
  • Generate custom character designs from a pose image

Model notes

  • The description lists 800 million parameters and the summary calls it 8B
  • Input length is stated as 512 tokens
  • The project is open source under the Apache 2.0 license
  • ComfyUI and Diffusers compatibility is listed

Getting started

  • Download the Flex.2-preview model files
  • Install dependencies such as torch and diffusers
  • Import the libraries in Python and load the model
  • Prepare the input image and the control image

Pros & cons

✓ What we like

  • Inpainting and control inputs are built into the model
  • Works with ComfyUI and Diffusers workflows

! What to watch out for

  • The material states a 512 token input length
  • Getting started calls for downloading model files and installing torch and diffusers

FAQ

What license is stated?

The material states the OSI Apache 2.0 license for the open-source project.

What control inputs are listed?

The feature list names pose, line and depth inputs.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review