Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · HiDream-ai

HiDream-I1-Full

HiDream-I1-Full is a text-to-image foundation model from HiDream-ai. According to the model card, it has 17B parameters, and captured Safetensors metadata reports about 17.11B parameters with MIT licensing.

Publisher
HiDream-ai
Task
text-to-image
Model type
Unknown
License
mit
Library
diffusers
Publication status
Accepted · not indexed

Model overview

HiDream-I1-Full is published by HiDream-ai as a text-to-image model. Captured Safetensors metadata reports 17,105,733,184 parameters, or about 17.11B. According to the model card, HiDream-I1 is an open-source image generative foundation model, with this Full variant run through Diffusers.

Recorded capabilities

17B scale

According to the model card, HiDream-I1 has 17B parameters; captured Safetensors metadata reports 17,105,733,184 parameters.

Reported benchmark tables

According to the model card, published tables cover DPG-Bench, GenEval, and HPSv2.1 comparisons against models including PixArt-alpha, SDXL, DALL-E 3, Flux.1-dev, and SD3 variants.

Publisher prompt-following claim

According to the model card, the model achieves industry-leading GenEval and DPG scores among open-source models; this is a publisher claim, not an Ethen verification.

MIT transformer licensing with component terms

According to the model card, Transformer weights use MIT licensing, while the VAE comes from FLUX.1 schnell and text encoders include google/t5-v1_1-xxl and Meta-Llama-3.1-8B-Instruct with their own terms.

Use cases in the source record

  • Text-to-image generation through the documented Diffusers HiDreamImagePipeline and GitHub inference scripts for full, dev, and fast variants, according to the model card.
  • Interactive generation experiments using the publisher-documented Gradio demo path, according to the model card.

Limitations and unknowns

  • The publisher's state-of-the-art and best-in-class quality claims are model-card claims and have not been independently verified by Ethen.
  • Reported DPG-Bench, GenEval, and HPSv2.1 figures are publisher-supplied tables without independent Ethen verification in this record.
  • Component licensing is mixed: MIT for Transformers, Apache-2.0 sources for VAE and T5, and Llama 3.1 Community License for the Llama encoder, according to the model card.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: HiDream-ai/HiDream-I1-Full

Captured: Unknown. Processed: 2026-09-07T19:35:06.328882+00:00.

HiDream-I1 is a new open-source image generative foundation model with 17B parameters that achieves state-of-the-art image generation quality within seconds. For more features and to experience the full capabilities of our product, please visit https://vivago.ai/ . Project Updates 🌟 July 16, 2025 : We've open-sourced the updated image editing model HiDream-E1.1 . 📝 May 28, 2025 : We've released our technical report HiDream-I1: A High-Efficient Image Generative Foundation Model with Sparse Diffusion Transformer . 🚀 April 28, 2025 : We've open-sourced the instruction-based-image-editing model HiDream-E1-Full . Experience at https:/…

F001F002F003F004F005F006F007F008F009F013F014F015F016F017F018F019F020