Tencent released the Hy Image3.5 preview model on September 22. The new iteration supports text-to-image, image-to-image, and multi-turn dialogue generation. According to internal blind tests conducted by hundreds of designers, the model shows a 30% performance improvement over Hy Image3.0, matching Seedream 5.0 pro and outperforming Nano-Banana Pro and Qwen-Image-3.0 Pro.
The model is now integrated into several Tencent platforms, including Yuanbao, ima, Miora, WorkRally, OnSolo, and WorkBuddy. Users of Miora, WorkRally, and OnSolo receive a two-week free trial period. Yuanbao and ima offer free access to specific features, such as portrait editing, smart image retouching, and knowledge-based infographics.
Regarding technical capabilities, Hy Image3.5 preview allows users to upload up to 5 reference images at once and generates outputs with a maximum resolution of 2K. The tool is designed for various professional applications, including commercial posters, storyboarding for television series, e-commerce assets, and daily image editing tasks.
Tencent Cloud has set the pricing for 2K image generation at $0.024 per image for the international version, while the domestic rate is ¥0.15 per image. The company charges only for the final output, with no additional fees for the input process.
