• Hunyuan is a multimodal large language model developed by Tencent.
• It is similar in power and capabilities to GPT-3, OpenAI’s core model as of 2022.
• Hunyuan has been seamlessly integrated into Tencent’s extensive ecosystem

Tencent, the prominent Chinese technology corporation, has unveiled its latest artificial intelligence system called “Hunyuan,” which is a multimodal large language model similar to OpenAI’s ChatGPT.

This announcement was made at the Global Digital Ecosystem Summit in Shenzhen on September 7th.

Hunyuan signifies Tencent’s venture into the realm of “foundational models,” designed as a framework to support various AI applications and interfaces. Notably, this AI system has been seamlessly integrated into Tencent’s extensive ecosystem, connecting with Tencent Cloud, Tencent Marketing Solutions, Tencent Games, Tencent fintech services, Tencent Meeting, Tencent Docs, Weixin Search, and QQ Browser.

This launch comes at a time when U.S.-China relations remain strained, exacerbated by an export ban on specific computer chips, including those crucial for AI system development and training. While diplomatic discussions continue, no resolution or timeline for lifting the ban has been established.

Tencent, on its part, asserts that the introduction of Hunyuan underscores its commitment to fostering open collaboration within the ecosystem. It aims to provide high-quality model services for domestic businesses while enabling international enterprises to access the Chinese market through Tencent’s offerings.

It’s important to note that ChatGPT, a counterpart developed by OpenAI, is not currently available in China. Although Hunyuan isn’t the first multimodal large language model in the Chinese market, its impact could be substantial. Tencent Holdings, as China’s largest technology company, has a widespread presence, particularly due to the ubiquity of WeChat and its affiliated applications. This positions Hunyuan to potentially penetrate the domestic market in ways that ChatGPT, developed by OpenAI and Microsoft, may struggle to replicate in the Western market.

Tencent claims that Hunyuan boasts similar power and capabilities to GPT-3, OpenAI’s core model as of 2022, in terms of raw numbers. With 100 billion parameters and 2 trillion tokens, this large language model stands among the most powerful in the world. However, it’s important to consider that parameter and token counts alone do not fully indicate the capabilities of a multimodal AI system. Hunyuan’s advantage lies in its extensive training on a massive corpus of Chinese language text, theoretically giving it an edge when operating in the Chinese language environment compared to models primarily trained on non-Chinese texts.

Also Read

Google’s new policy on political ads aims to prevent AI-driven deception

Google will require political ads to disclose the use of AI-generated images,...