What can I use P-H3-Image-Comfy for?

Type a one-line request describing a scene into the text-to-image workflow, and you get an 8-megapixel image. Load a photo and optional reference images into the edit workflow, and you get an edited version of the photo. Type a request and choose the 16:9 size in the workflow, and you get a 3872 × 2176 image.

What do I need to use P-H3-Image-Comfy?

A graphics card with about 21 GB to 56 GB of memory (VRAM), depending on the file set, and ComfyUI updated to the latest GitHub version.

P-H3-Image-Comfy: Make and edit 8-megapixel images in ComfyUI

Model details

Original model PrunaAI/P-H3-Image-Comfy

Make/edit images

Input Text Output Image

Try the browser demo Follow the model card setup

Ways to use it Browser demo · Run locally

Family context

Role in this family: Smaller quantized version.

PrunaAI/p-h3-image

What it does

P-H3-Image-Comfy is a set of ComfyUI files from Pruna AI for P-H3-Image, an image model based on MiniMax-H3. It turns text into 8-megapixel images in 8 steps and edits pictures with up to 8 input images. A helper model called the prompt upsampler first expands a short request into a fuller prompt.

How to use it

Try the model first in the web demo linked from the model card. To run it yourself, update ComfyUI, download one file per component into its models folder, open a workflow from the workflows folder, type a short request and run.

Common use cases

Type a one-line request describing a scene into the text-to-image workflow, and you get an 8-megapixel image.
Load a photo and optional reference images into the edit workflow, and you get an edited version of the photo.
Type a request and choose the 16:9 size in the workflow, and you get a 3872 × 2176 image.

Can I use this?

Setup details on the source page

Requirements It runs on your own computer through ComfyUI, and the smaller files need NVIDIA Ampere or newer or AMD RDNA3/4 graphics cards. Downloads run from 20 GB to 55 GB, plus the prompt upsampler. ComfyUI's desktop app and Comfy Cloud follow later stable releases.

Caveat Download and memory figures are the creators' own. Licensed under the MiniMax-H3 Community License Agreement, a custom license; read the License file linked on the model page for its terms.

Worth knowing The smallest w4a8 files change more details and sometimes the composition. Without the prompt upsampler, short requests render much worse than full, specific prompts.

Why it's here

236 people have liked it on Hugging Face.
It was downloaded 625,136 times in the last 30 days.

Numbers as of 8 November 2026, from BuildTube’s latest daily refresh.

Behind it

View model page

BuildTube has not run or verified this project. Everything above is written from what the creator published.

Made this?

Quick questions

What is P-H3-Image-Comfy?: P-H3-Image-Comfy is a set of ComfyUI files from Pruna AI for P-H3-Image, an image model based on MiniMax-H3.
What can I use P-H3-Image-Comfy for?: Type a one-line request describing a scene into the text-to-image workflow, and you get an 8-megapixel image. Load a photo and optional reference images into the edit workflow, and you get an edited version of the photo. Type a request and choose the 16:9 size in the workflow, and you get a 3872 × 2176 image.
Can I run P-H3-Image-Comfy on my computer?: Downloads run from 20 GB to 55 GB, plus the prompt upsampler.
What do I need to use P-H3-Image-Comfy?: A graphics card with about 21 GB to 56 GB of memory (VRAM), depending on the file set, and ComfyUI updated to the latest GitHub version.
What license does P-H3-Image-Comfy use?: Licensed under the MiniMax-H3 Community License Agreement, a custom license; read the License file linked on the model page for its terms.

More from this week

Idol Turn one picture of an object into a 3D Gaussian splat
Idol turns one photo, render or generated image of an object into a 3D Gaussian splat file.
Runs in a terminalAdded 2 Nov
agent-desktop Let AI agents read and click desktop apps
For developers building computer-use agents: agent-desktop lets an AI agent read and operate desktop apps on Mac, Windows or Linux from a command line.
+13,858 stars this weekAdded 3 Nov
Auto-Ontology Give AI agents checked business meanings for your database
Help AI agents answer questions about company databases using reviewed definitions and SQL, with each answer showing its steps.
+9,044 stars this weekNeeds a local setupAdded 4 Nov

Related projects

Qwen-Image-2.1-viggle-turbo Make images about 5x faster with a distilled Qwen model
Same topic: comfyui
Needs a model setupAdded 1 Oct
flux-3-image-klein-comfy Make and edit images in ComfyUI (5.6B)
Same topic: diffusion-single-file
Added 22 Oct
Qwen-Image-2.1-GGUF Run Qwen’s image model on your laptop, compressed
Same topic: text-to-image
Needs a model setupAdded 1 Oct
flux-3-image-klein Generate and edit images (5.6B)
Same topic: image-to-image
Needs a model setupAdded 20 Oct
Qwen-Image-2.1-Turbo Generate and edit images with an eight-step Qwen model
Same topic: text-to-image
Needs a model setupAdded 10 Oct
Qwen-Image-2.1 Generate and edit images, including transparent layers, from text
Same topic: text-to-image
Needs a model setupAdded 25 Sept

Explore by topic

Post this

For makers

Made this? Show it’s on BuildTube — add the badge to your README or embed the card on your site.

README badge (Markdown)

[![Featured on BuildTube](https://buildtube.app/badge/hf-prunaai-p-h3-image-comfy.svg)](https://buildtube.app/p/hf-prunaai-p-h3-image-comfy/)

Embed card (HTML)

<iframe src="https://buildtube.app/embed/hf-prunaai-p-h3-image-comfy/" width="560" height="240" style="border:0;max-width:100%" loading="lazy" title="P-H3-Image-Comfy on BuildTube"></iframe>