Home
Tencent Accelerates AI Push With Doubled-Size Hunyuan Hy4, Morgan Stanley Lifts Target to HKD 550
Tencent has officially unveiled the preview of HuanYuan Hy4, accelerating its timeline ahead of Morgan Stanley’s September forecast and fulfilling the company’s Q2 earnings commitment to launch the model within the year. Since appointing Yao Shunyu as Chief AI Scientist in Q4 2025, Tencent has significantly accelerated its model iteration. Following the Hy3 preview in April and the official Hy3 release in July, the early arrival of the Hy4 preview suggests the official launch—originally slated for Q4—may occur even sooner than anticipated.
Technically, the HuanYuan Hy4 preview utilizes a Mixture of Experts (MoE) architecture, marking a substantial leap in scale and long-context processing. The model boasts 770 billion total parameters, double Hy3’s 295 billion, with active parameters rising to 49 billion (up from 21 billion). Its context window now supports up to 1 million tokens, vastly outperforming Hy3’s 256,000-token limit.

Driven by high-quality training data and optimized software-hardware integration, the Hy4 preview has reached industry-leading performance levels. In Tencent’s internal double-blind evaluations, it achieved an average score of 2.99/4, edging out comparable products, and ranked highly in third-party coding benchmarks. Beyond standard tests, the model delivers tangible productivity gains in software development, office workflows, game creation, and research. These capabilities align closely with Tencent’s core businesses in gaming, advertising, and cloud services, positioning them for sustained commercial growth.
Leveraging Tencent’s AI advancements and strong fundamentals, Morgan Stanley maintains an “overweight” rating with a target price of 550 HKD. The firm projects steady revenue and profitability growth from 2026 to 2028, anticipating that Tencent’s valuation metrics and return on equity will remain highly competitive in the market.
Related article
Pearson’s Dave Treat: AI Should Guide Learning, Not Cheat
Pearson’s Chief Technology Officer, Dave TreatOn academic integrity, AI adoption, and the critical skills that matter as agentic AI transforms educationDave Treat joined Pearson nearly two years ago as Chief Technology Officer, later assuming the rol
OpenAI Launches GPT-5.6-Cyber, Boosting Zero-Day Detection with AI Automation
OpenAI has expanded its Daybreak cybersecurity initiative, introducing a two-tier access system and launching GPT-5.6-Cyber, a model purpose-built for security operations. This move addresses the growing reality that threat actors are leveraging AI t
Broadcom predicts Anthropic and OpenAI will surpass Google as top AI custom chip customers
During the third quarter of fiscal year 2026, Broadcom President Chen Fuyang delivered a pivotal forecast on the earnings call, noting that intensifying competition among major AI model developers is reshaping Broadcom’s core customer base. He projec
Related Special Topic Recommendations
Comments (0)
0/500
Tencent has officially unveiled the preview of HuanYuan Hy4, accelerating its timeline ahead of Morgan Stanley’s September forecast and fulfilling the company’s Q2 earnings commitment to launch the model within the year. Since appointing Yao Shunyu as Chief AI Scientist in Q4 2025, Tencent has significantly accelerated its model iteration. Following the Hy3 preview in April and the official Hy3 release in July, the early arrival of the Hy4 preview suggests the official launch—originally slated for Q4—may occur even sooner than anticipated.
Technically, the HuanYuan Hy4 preview utilizes a Mixture of Experts (MoE) architecture, marking a substantial leap in scale and long-context processing. The model boasts 770 billion total parameters, double Hy3’s 295 billion, with active parameters rising to 49 billion (up from 21 billion). Its context window now supports up to 1 million tokens, vastly outperforming Hy3’s 256,000-token limit.

Driven by high-quality training data and optimized software-hardware integration, the Hy4 preview has reached industry-leading performance levels. In Tencent’s internal double-blind evaluations, it achieved an average score of 2.99/4, edging out comparable products, and ranked highly in third-party coding benchmarks. Beyond standard tests, the model delivers tangible productivity gains in software development, office workflows, game creation, and research. These capabilities align closely with Tencent’s core businesses in gaming, advertising, and cloud services, positioning them for sustained commercial growth.
Leveraging Tencent’s AI advancements and strong fundamentals, Morgan Stanley maintains an “overweight” rating with a target price of 550 HKD. The firm projects steady revenue and profitability growth from 2026 to 2028, anticipating that Tencent’s valuation metrics and return on equity will remain highly competitive in the market.
Pearson’s Dave Treat: AI Should Guide Learning, Not Cheat
Pearson’s Chief Technology Officer, Dave TreatOn academic integrity, AI adoption, and the critical skills that matter as agentic AI transforms educationDave Treat joined Pearson nearly two years ago as Chief Technology Officer, later assuming the rol
OpenAI Launches GPT-5.6-Cyber, Boosting Zero-Day Detection with AI Automation
OpenAI has expanded its Daybreak cybersecurity initiative, introducing a two-tier access system and launching GPT-5.6-Cyber, a model purpose-built for security operations. This move addresses the growing reality that threat actors are leveraging AI t
Broadcom predicts Anthropic and OpenAI will surpass Google as top AI custom chip customers
During the third quarter of fiscal year 2026, Broadcom President Chen Fuyang delivered a pivotal forecast on the earnings call, noting that intensifying competition among major AI model developers is reshaping Broadcom’s core customer base. He projec











