# HiDream Future Launches Commercial Image Generation Model HiDream-O1-Image-1.5

- Category: Flash (快讯)
- Published: 2026-06-11T11:58:00.000Z
- Source date: 2026-06-11
- HTML: https://www.88z.ai/en/news/flash-20260611-18ab77e9d91e7970
- Markdown: https://www.88z.ai/en/news/flash-20260611-18ab77e9d91e7970.md
- Other language: https://www.88z.ai/news/flash-20260611-18ab77e9d91e7970.md

HiDream Future has released the commercial image generation model HiDream-O1-Image-1.5. Ranking third globally and first in China on the Artificial Analysis text-to-image leaderboard with an ELO of 1265, it trails only OpenAI's GPT Image series while surpassing major models from Google, NVIDIA, and ByteDance. Built on the native multimodal UiT architecture, it excels in semantic adherence, text rendering, complex layout, and multi-subject consistency.

## At a glance

- HiDream-O1-Image-1.5 ranks third globally and first in China on the Artificial Analysis leaderboard
- Built on UiT architecture, it excels in semantic adherence and text rendering

## The update

HiDream Future has officially launched HiDream-O1-Image-1.5, a commercial image generation model. According to data from the Artificial Analysis text-to-image leaderboard, the model achieved a global rank of third and first in China with an ELO score of 1265. In this ranking, it trails only OpenAI's GPT Image series while surpassing mainstream models from competitors such as Google, NVIDIA, and ByteDance. Technically, HiDream-O1-Image-1.5 is built on the native multimodal UiT architecture. This framework enables outstanding performance in several key areas, specifically demonstrating strong capabilities in semantic adherence, text rendering quality, complex layout handling, and maintaining consistency across multiple subjects.

## Sources

- [HiDream智象未来](https://mp.weixin.qq.com/s/I0prC7MjxYwcFWh3InBIEg) (2026-06-11)

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
