P-H3-Image
Pruna AI's 33B text-to-image and image-editing model derived from MiniMax-H3, with a ComfyUI variant.
This entry is marked for review. Check the original sources before relying on its details.
Overview
P-H3-Image is a 33B-parameter text-to-image and image-editing model from Pruna AI, based on MiniMax-H3 and published on Hugging Face around November 3, 2026, with a ComfyUI variant (P-H3-Image-Comfy) and a ZeroGPU demo Space. A release tracker reports Pruna says it generates 4 MP images about five times faster than MiniMax-H3 on a B200; that figure is vendor-reported via a secondary source and is not stated on the model page. The license is not stated on the page checked.
This is a sourced directory entry. Features and capabilities are reported in the linked sources; this entry does not document independent Aymorphic testing.
Key features
- 33B parameters, text-to-image and image editing
- Derived from MiniMax-H3
- ComfyUI variant published alongside (P-H3-Image-Comfy)
- Reported 5x faster 4 MP generation than MiniMax-H3 on a B200 (vendor claim, secondary source)
- Hugging Face ZeroGPU demo Space
Use cases
- Fast high-resolution image generation in ComfyUI workflows
- Prompt-based image editing
- Benchmarking inference-optimized variants of a base image model
Pricing
Weights on Hugging Face; license unknown; Pruna offers an API with unpublished pricing.
Pricing reflects the source record and may change. Confirm current terms with the provider.
Dates and sources
- Released
- 3 Nov 2026
- Record updated
- 6 Nov 2026
- Record last checked
- 6 Nov 2026