zenlm/zen3-image-sdxl

Zen3 Image Sdxl

SDXL-compatible variant of Zen3 Image for high-quality diffusion-based generation.

Overview

Built on Zen MoDE (Mixture of Distilled Experts) architecture with 12B parameters.

Developed by Hanzo AI and the Zoo Labs Foundation.

Quick Start

from diffusers import AutoPipelineForText2Image
import torch

model_id = "zenlm/zen3-image-sdxl"
pipe = AutoPipelineForText2Image.from_pretrained(model_id, torch_dtype=torch.bfloat16)
pipe = pipe.to("cuda")

image = pipe("A serene mountain landscape at sunset, photorealistic").images[0]
image.save("output.png")

API Access

from openai import OpenAI

client = OpenAI(base_url="https://api.hanzo.ai/v1", api_key="your-api-key")
response = client.images.generate(
    model="zen3-image-sdxl",
    prompt="A serene mountain landscape at sunset",
    size="1024px",
)
print(response.data[0].url)

Model Details

Attribute	Value
Parameters	12B
Architecture	Zen MoDE
Max Resolution	1024px
License	Apache 2.0

License

Apache 2.0