Free News Reader

Tencent Launches Advanced AI Image Model Amidst Rival Innovations

Free News Reader  ·  September 21, 2026

AI-generated context summary requested by a Free News Reader user. Sourced via Gemini from publicly available information — no paywalled content was accessed.

You hit a paywall. Here’s the context on this topic based on publicly available information. We did not access any paywalled content. View original article.

Tencent Launches Advanced AI Image Model Amidst Rival Innovations

  • Tencent Holdings Ltd. released its Hy Image 3.5 Preview model on Tuesday, September 22, 2026, which boasts a 30% performance improvement over its predecessor.
  • This launch coincided with Alibaba Group Holding Ltd.'s Apsara Conference, where Alibaba announced its new Zhenwu V900 AI chip and plans for a Qwen model with up to 10 trillion parameters.

Full Summary — powered by AI

Tencent Holdings Ltd. officially launched its Hy Image 3.5 Preview on Tuesday, September 22, 2026. This new image-generation model is designed to enhance performance by approximately 30% compared to its previous version, Hy Image 3.0. The model facilitates text-to-image and image-to-image generation, supports multi-turn conversational editing, and can produce images with a maximum resolution of 4K. Tencent is integrating Hy Image 3.5 Preview into various products, including its Yuanbao chatbot and existing film-editing and design tools. The company states that internal blind testing with hundreds of designers showed the model’s outputs were comparable to ByteDance’s Seedream 5.0 Pro.

The release of Tencent’s new AI image model occurred on the same day that rival Alibaba Group Holding Ltd. commenced its annual Apsara Conference in Hangzhou. At the conference, Alibaba unveiled its latest AI chip, the Zhenwu V900, which is scheduled for mass production and commercial release in the first quarter of 2027 and is reported to offer three times the performance of its predecessor. Additionally, Alibaba announced plans to develop a new Qwen AI model with an ambitious scale of 5 trillion to 10 trillion parameters, significantly larger than its current flagship Qwen 3.8 Max, which has 2.4 trillion parameters. Alibaba Group Chairman Joe Tsai emphasized that AI is transitioning from theoretical exploration to practical, large-scale deployment.

This simultaneous timing highlights the intensifying competition among Chinese technology giants in the generative AI space, with companies like Tencent, Alibaba, and ByteDance actively advancing their AI capabilities across various applications. ByteDance, known for TikTok, has its own advanced AI image generation model, Seedream 5.0 Pro, which features a layer-based editing system and excels in producing photorealistic images and handling text-dense outputs.