hidream-i1-full-t2i

HiDream by Zhipu AI is the world's first open-access image and video generation model based on the Diffusion Transformer (DiT) architecture. As a core product of Zhipu AI's multi-modal generative large model series, it has over 13 billion parameters and powerful capabilities for generating text, images, and videos in multiple modalities.

The HiDream series models have been downloaded more than 600,000 times cumulatively and integrated into mainstream tools such as the Diffusers library, ComfyUI, and Recraft, demonstrating significant international influence. In April this year, Zhipu AI released HiDream-E1, the world's first conversational interactive editing model. Users can complete real-time image editing through natural language instructions, truly achieving "what you see is what you get." Within 24 hours of its release, it quickly topped international rankings.

As part of Zhipu AI's toolchain built around the "agent" concept, HiDream not only innovates at the technical level but also delves deeper into application fields, providing revolutionary solutions for image generation and video creation, continuously advancing the forefront of AI technology development.

302.AI Release date: 2025-04-12