P-Image-OSS
Adapts the MiniMax H3 backbone to still-image generation and multi-reference edits.
- Arena score
- —Not enough votes yet
- Starting price
- —Not published
- Largest size
- —Not published
- References
- Yesinput images for editing
How it rates and what it costs
P-Image-OSS does not have enough arena votes for a quality score yet.
The details
- Lab
- Pruna
- Task
- Image generation and editing
- Published capabilities
- Adapts the MiniMax H3 backbone to still-image generation and multi-reference edits.
- Access
- Public weights, ComfyUI workflow and Hugging Face demo
- Operating limits
- 2–16 megapixels and up to nine input images. The smallest published build is 20 GB.
- Sources
- Publisher documentation · Model weights
- Type
- Image model
- Released
- Nov 3, 2026
- Inputs
- Text prompts, Reference images
- Output
- Images
- Can do
- Text to image, Image editing
- Weights
- Open weights
- Availability
- Not on the API hosts we track
More image models from Pruna
About P-Image-OSS
Who makes P-Image-OSS?
P-Image-OSS is an image model from Pruna, released Nov 3, 2026.
What can P-Image-OSS do?
P-Image-OSS accepts text prompts and reference images and produces images. In practice: text to image and image editing.
Is P-Image-OSS open source?
Yes. Pruna publishes the weights, so you can run it on your own hardware.
Where can I use P-Image-OSS?
Public weights, ComfyUI workflow and Hugging Face demo
What are the alternatives to P-Image-OSS?
The closest image models from other labs, by release date, are MiniMax H3 Image (MiniMax, Nov 10, 2026), GPT Image 2.6 Flare (OpenAI, Nov 4, 2026), SenseNova U1.5 TailSteer (SenseTime, Nov 4, 2026) and MAI-Image-2.7 (Microsoft, Nov 17, 2026).