Tencent shares surge after new AI image model launch

Tencent unveiled its most advanced image-generation model on Tuesday, sending its Hong Kong-listed shares up as much as 7% as investors drew fresh parallels between the Chinese tech giant and Meta ↗ 1.30% in the race to build AI-powered super-apps.

A New Model Amid Fierce Competition

The Hy Image 3.5 Preview, released the same day rival Alibaba ↗ 0.79% kicked off its annual Apsara AI conference, supports text-to-image generation, image-to-image generation, and multi-turn editing at resolutions up to 4K. Tencent said the model delivers roughly a 30% performance improvement over its predecessor. Generating a 2K image through the Tencent Cloud API costs just 0.15 yuan — about 1.5 U.S. cents — with no charge for failed outputs.news.aibase+3

In internal blind tests conducted by hundreds of Tencent's in-house designers, the model matched ByteDance's Seedream 5.0 Pro and slightly outperformed Alphabet's ↗ 0.35% Nano Banana Pro and Alibaba's Qwen-Image-3.0 Pro, according to Bloomberg. Tencent did not provide independent metrics to quantify those claims.businesstimes+1

WeChat, Meta Comparisons Fuel Rally

The stock move reflected more than just a product launch. Meta's personal AI agent Muse surged to the top of U.S. app store charts after its September 8 release, drawing roughly 730,000 downloads in five days and reigniting investor appetite for platforms that can embed AI into large social ecosystems. "Some investors are drawing parallels between Tencent and Meta, particularly given WeChat's unique social ecosystem and its potential to support a personal AI assistant at scale," said Gary Tan, portfolio manager at Allspring Global Investments.kucoin+2

Tencent has not yet released a direct rival to Muse, nor disclosed a commercialization timeline, meaning the rally largely reflects a market repricing of WeChat's AI potential rather than new revenue.kucoin

Talent Push and Strategic Shift

Tencent has historically lagged Big Tech peers on large language models but has been retooling under new leadership. Former OpenAI researcher Yao Shunyu, now chief AI scientist, overhauled the company's AI infrastructure and has argued against chasing benchmark leaderboards in favor of product integration. In July, another former OpenAI researcher, Tian Yonglong, joined to lead vision-language models under the Hunyuan team. The Hy Image 3.5 Preview is already rolling out across Tencent products including the Yuanbao chatbot and film-editing tools, with demand for image generation rising more than 50% during gray-scale testing.news.aibase+1