Home
Tencent Unveils Hunyuan Image 3.5: Enhanced Text Drawing and Editing with 2K Images at Just 1.5 Cents
Tencent Huan Yuan launched the preview of Image 3.5 (Hy Image3.5preview) on September 22, delivering substantial upgrades in image quality, layout precision, and photo editing while drastically cutting costs under the tagline "affordable and accessible."

Official details indicate a 30% performance boost over the previous generation. The model now renders text with greater accuracy, eliminating common artifacts like distorted limbs. Layout understanding has also improved, allowing for more precise visual arrangements. For photo editing, users can input up to five reference images to blend styles or elements, with outputs reaching 2K resolution.

Pricing remains highly competitive: generating a 2K image via the Tencent Cloud API costs just 0.15 yuan, with billing applied only to successful outputs. In internal blind tests, the model matched Seedream5.0pro and outperformed two other competitors.
Hy Image3.5preview supports diverse applications, including daily photo editing, poster design, and storyboard creation. It is currently integrated into Tencent ecosystem products like Yuanbao and ima, with direct API access available for developers.

During gray-scale testing, drawing demand surged by over 50%. One film studio reported that batch processing with Huan Yuan Image 3.5 significantly reduced pre-production time, boosting overall workflow efficiency.
Tencent Huan Yuan plans to continue refining image generation, enhancing performance in complex scenarios while preserving cost-effectiveness.
Hy Blog:
https://hy.tencent.com/research/hy-image35-preview
Model Experience:
https://aistudio.tencent.com/
Official Prompt Manual:
https://page.weixin.qq.com/smartpage/p/b1_ALAAAwZ2ACcbt4quDAMT0i1obS0
Related article
China Replaces All K-12 Textbooks, Lowers Difficulty and Integrates AI into Classrooms
This week marks the official start of the autumn semester for primary and secondary schools nationwide, bringing a series of new adjustments to compulsory education. Textbooks for three core subjects, previously piloted in phases starting in autumn 2
OpenAI backs Sam Altman’s Merge Labs brain-computer interface venture
In a move that deepens the intersection of AI and human biology, OpenAI has invested in Merge Labs, a brain-computer interface startup founded by CEO Sam Altman.Emerging from stealth on Thursday, Merge Labs announced an undisclosed seed round. Source
Similarweb Report: Google AI Overview Appearance Rate Rises to 43% in One Year, AI Search Becomes a Main Entry Point
Recent Similarweb data reveals that AI search is fundamentally reshaping user behavior. Google’s AI Overviews now appear in 43% of results, up from just 15% a year ago, establishing AI-driven discovery as the standard user experience.Initially a supp
Related Special Topic Recommendations
Comments (0)
0/500
Tencent Huan Yuan launched the preview of Image 3.5 (Hy Image3.5preview) on September 22, delivering substantial upgrades in image quality, layout precision, and photo editing while drastically cutting costs under the tagline "affordable and accessible."

Official details indicate a 30% performance boost over the previous generation. The model now renders text with greater accuracy, eliminating common artifacts like distorted limbs. Layout understanding has also improved, allowing for more precise visual arrangements. For photo editing, users can input up to five reference images to blend styles or elements, with outputs reaching 2K resolution.

Pricing remains highly competitive: generating a 2K image via the Tencent Cloud API costs just 0.15 yuan, with billing applied only to successful outputs. In internal blind tests, the model matched Seedream5.0pro and outperformed two other competitors.
Hy Image3.5preview supports diverse applications, including daily photo editing, poster design, and storyboard creation. It is currently integrated into Tencent ecosystem products like Yuanbao and ima, with direct API access available for developers.

During gray-scale testing, drawing demand surged by over 50%. One film studio reported that batch processing with Huan Yuan Image 3.5 significantly reduced pre-production time, boosting overall workflow efficiency.
Tencent Huan Yuan plans to continue refining image generation, enhancing performance in complex scenarios while preserving cost-effectiveness.
Hy Blog:
https://hy.tencent.com/research/hy-image35-preview
Model Experience:
https://aistudio.tencent.com/
Official Prompt Manual:
https://page.weixin.qq.com/smartpage/p/b1_ALAAAwZ2ACcbt4quDAMT0i1obS0
China Replaces All K-12 Textbooks, Lowers Difficulty and Integrates AI into Classrooms
This week marks the official start of the autumn semester for primary and secondary schools nationwide, bringing a series of new adjustments to compulsory education. Textbooks for three core subjects, previously piloted in phases starting in autumn 2
OpenAI backs Sam Altman’s Merge Labs brain-computer interface venture
In a move that deepens the intersection of AI and human biology, OpenAI has invested in Merge Labs, a brain-computer interface startup founded by CEO Sam Altman.Emerging from stealth on Thursday, Merge Labs announced an undisclosed seed round. Source
Similarweb Report: Google AI Overview Appearance Rate Rises to 43% in One Year, AI Search Becomes a Main Entry Point
Recent Similarweb data reveals that AI search is fundamentally reshaping user behavior. Google’s AI Overviews now appear in 43% of results, up from just 15% a year ago, establishing AI-driven discovery as the standard user experience.Initially a supp











