Tencent Hunyuan releases Hunyuan Image 3.5 preview version, capable of meeting the needs of professional-grade aesthetic scenarios.

On September 22, Tencent Hunyuan officially released the Hunyuan Image 3.5 preview version (Hy Image3.5 preview). This model supports text-to-image and image-to-image generation, allows up to 5 images to be uploaded as references in a single session, and supports output in multiple size ratios, achieving up to 2K resolution output. The model supports multi-turn dialogue-based creation and can perform continuous creation and editing based on historical context. According to the introduction, Hy Image3.5 preview continues the design approach of the Hy Image3.0 series, with significant improvements in text rendering and layout, realism and stylistic expressiveness, and editing consistency. It provides outstanding image generation capabilities for daily creation and professional visual production at a highly attractive cost-performance ratio, with comprehensive capabilities significantly improved compared to the previous generation model. Hy Image3.5 preview has undergone a systematic upgrade centered on "from intent understanding to visual expression." It can serve everyday entertainment and learning scenarios such as daily photo editing, portrait beautification, and knowledge cards, as well as professional-level scenarios including commercial posters, episode storyboards, film and television materials, UI design, creative illustration, brand visuals, and e-commerce materials, enabling generated results to move from "looking good" to meeting "professional-grade aesthetic scenario needs." This is a highly cost-effective professional-grade image generation model. After GSB blind testing by hundreds of internal professional designers, the model's capabilities improved by approximately 30% compared to Hy Image3.0. Through deep collaboration with products such as Yuanbao, WorkRally, OnSolo, Miora, WorkBuddy, and ima, it greatly enhances effectiveness and capabilities for life, learning, and productivity scenarios. In terms of pricing, Hunyuan further improved inference resource utilization to provide a foundation for highly cost-effective image generation. Hy Image3.5 preview's Tencent Cloud API 2K image pricing is 0.15 yuan per image, making its cost-performance highly competitive; meanwhile, Hy Image3.5 preview only charges for output images. Currently, the model has been launched on multiple native AI applications including Tencent AI assistant Yuanbao, professional film and television intelligence platform WorkRally, professional series and game content production tool OnSolo, creative agent Miora, office agent workspace WorkBuddy, and AI knowledge manager ima. The API capabilities have also been simultaneously launched on Tencent Cloud TokenHub, Tencent Cloud Media Processing Service MPS, and Cloud Video on Demand VOD.