HiDream-O1-Image-1.5 - A commercial image generation model launched by Zhixiang Future.

What is HiDream-O1-Image-1.5?

HiDream-O1-Image-1.5 is a commercial image generation model launched by Zhixiang Future, based on a native full-modal UiT architecture. In the Artificial Analysis raw image leaderboard, it ranked third globally and first in China with an ELO score of 1265, surpassing Google Nano Banana 2 and ByteDance's Seedream 4.0. The model possesses photographic-grade portrait and animal modeling capabilities, accurate text rendering, and multi-subject consistency, targeting commercial scenarios such as advertising, brand design, e-commerce visuals, and film storyboarding. This signifies Zhixiang Future's firm position among the world's leading companies in the field of visual generation.

Main functions of HiDream-O1-Image-1.5

• Portrait photography generationIt supports magical lighting and shadows, two-player interaction, and character close-ups, and presents natural effects in terms of skin texture, clothing texture, body relationships, and environmental blurring.
• Animals and their natural environment: Detailed modeling of animal structure, fur texture, dynamic performance, and complex lighting, underwater refraction, etc.
• Text rendering and layoutIt possesses accurate text generation capabilities and complex typesetting capabilities.
• Multi-subject consistencySupports the coordinated generation of multiple characters and elements, and visual storytelling.
• Storyboarding and Scene ConstructionSupports complex compositions such as film and television storyboards and wide-angle/low-angle shots.

Technical Principles of HiDream-O1-Image-1.5

• Native full-modal UiT architectureThe model is based on Zhixiang Future's self-developed Unified Transformer (UiT) native multimodal architecture. The architecture uses a unified pixel-level native representation to process multimodal information, avoiding information loss caused by modality conversion in traditional multimodal models, and enabling text, image and other data to be understood and generated in a unified space.
• From open-source verification to commercial productionThe model continues the technical roadmap of the open-source version HiDream-O1-Image-Dev-2604, advancing the UIT architecture from technical verification to production verification. Building upon the pixel-level native full-modal capabilities already verified in the open-source version, the commercial version is enhanced and optimized for high-demand commercial scenarios such as advertising and marketing, brand design, and e-commerce visuals, transforming the advantages of the underlying architecture into a visual productivity tool.
• Comprehensive capability enhancement mechanismThe model achieved an ELO score of 1265 in an anonymous comparative evaluation of over 4000 samples by improving semantic compliance accuracy, stability in generating complex scenes, text rendering accuracy, and multi-subject consistency control. The core technology lies in end-to-end joint modeling of deep semantic understanding of text instructions and pixel-level image generation, ensuring the harmonious unity of complex compositions, spatial perspective, and visual narrative.