The article notes Tencent Cloud’s prior involvement with Pakistan’s Tapmad streaming platform to handle 7 million concurrent viewers, illustrating scalability, though this is context rather than a Hy3 milestone.

Hy3, with about 295 billion parameters, is not the largest but Tencent claims it delivers the best performance among its AI developments and focuses on practical, scalable workplace use.

Hy3 is accessible through WorkBuddy (free to users worldwide until August 31, 2026), Tencent Design Miora, and Tencent Cloud TokenHub, with API access for developers to integrate it into various tools.

Hy3 follows the Hy series’ 2023 reveal as Hunyuan, with prior performance gaps prompting an overhaul of R&D, data, and evaluation systems to support Hy3.

Hy3 has shown strong early adoption, delivering more API calls than its predecessor and ranking first on OpenRouter’s global LLM usage leaderboard within a week of launch, with OpenRouter pricing starting at $0.1288 per million input tokens and $0.5336 per million output tokens.

In internal Tencent testing, using Hy3 in WorkBuddy raised task success rate above 90% and reduced average task completion time by 34% versus the previous Hy model.

Industry observers view the global rollout as a stress test of Tencent’s revamped AI strategy and a key step toward expanding overseas user base and leadership in the generative AI race.

The expansion is part of a broader Chinese AI competition among Tencent, Alibaba, ByteDance, DeepSeek, and Zhipu AI, with Tencent aiming to grow an AI agent ecosystem centered on Hy3 through WorkBuddy and related tools.

Poshu Young, SVP of Tencent Cloud, emphasizes Hy3’s practical, reliable, and scalable deployment, helping enterprises move from pilots to large-scale deployments with enterprise-grade engineering and secure infrastructure.

Quotes from Poshu Yeung underscore the aim to move enterprises from experimentation to scalable, trusted AI deployments.

Tencent stresses enterprise-grade reliability, secure infrastructure, and seamless scaling from pilots to full deployment across cloud and product ecosystems.

Market context notes ongoing overseas interest in Chinese-origin open-source models, but enterprise adoption lags in some markets due to data privacy and security concerns; Tencent plans to address these with open access and open-source strategies.