Skip to content

PCM

A consistency model for high-resolution text-to-image

Visit official site

What PCM is

The model is described as a new generative model targeting the limits of Latent Consistency Model in text-conditioned high-resolution work. New strategies are used in both training and inference.

What you can do with it

  • Generate images that match a text description
  • Pair the model with SDXL for more steps
  • Keep video output stable in low-step ranges

What to watch out for

  • Base models are obtained before setup
  • Parameters are configured for each task
  • Training data is prepared by the user

Pros & cons

✓ What we like

  • Training and inference strategies are both listed
  • Step counts from 1 to 16 are examined

! What to watch out for

  • Base models such as Stable Diffusion are needed
  • Parameters and training data are configured by the user

FAQ

Which step counts are listed?

The material names 1, 2, 4, 8 and 16 steps.

Which base models?

Stable Diffusion and Stable Diffusion XL.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review