HiDream-O1-Image: 8B Pixel-Level Unified Transformer by HiDream-ai

HiDream-O1-Image is an 8B parameter natively unified image generative foundation model built on a Pixel-level Unified Transformer (UiT), supporting text-to-image, image editing, and subject-driven personalization at up to 2048×2048.

H

HiDream-O1-Image by HiDream-ai

Text-to-ImageImage Editing8BUiTOpen Source

Natively unified 8B image generative foundation model built on a Pixel-level Unified Transformer (UiT). No external VAE or disjoint text encoders — encodes raw pixels, text, and task conditions in a single shared token space. Supports text-to-image, long-text rendering, instruction editing, subject-driven personalization, and storyboard generation at up to 2048×2048.

Hugging Face (Upstream)HF Comfy-OrgGitHubPaper
DeveloperHiDream-ai
Release Date2026-05-08
ArchitecturePixel-level Unified Transformer (UiT) — 8B parameters
LicenseMIT
Output ResolutionUp to 2048×2048 native
Inference StepsFull: 50 steps; Dev (distilled): 28 steps
Quantized VariantsBF16, FP8 scaled, MXFP8
CapabilitiesText-to-Image, Image Editing, Subject-Driven Personalization, Long-Text Rendering, Storyboard Generation

Guides and workflows related to this model series.

No articles found.