China’s Tencent has introduced a new open-source AI model that can operate computers, interpret complex images, and tackle demanding engineering tasks. Called Hy4, the model marks Tencent’s latest attempt to compete with leading open-source AI systems.Hy4 adds visual understanding to Tencent’s Hy model family for the first time. It can process detailed images, interpret complicated charts, and solve mathematical problems involving visual information. The model also introduces computer-use capabilities, allowing it to interact with command-line tools and graphical interfaces.Tencent plans to publish Hy4’s model weights and system cards in the coming days. The company says it has also incorporated safeguards targeting cybersecurity and biological risks.Computer control gets smarterHy4 can operate professional software, including Blender and computer-aided design (CAD) applications. Tencent trained the model to handle complex workflows that require planning and multiple execution steps. These capabilities could help automate tasks that traditionally demand considerable manual input.The model also supports 3D scene creation, web development, and document production. Tencent reported an average evaluation score of 83.18 on its WorkBuddy platform. Hy4 achieved a 96.15 percent pass rate across evaluated tasks.It also ranked first in nine of 12 tested scenarios within mainland China. WorkBuddy integrates Hy4 with Tencent’s workplace collaboration software, WeCom. The integration lets the model access information across chats, emails, documents, calendars, and meetings.Hy4 can track project developments, identify unfinished assignments, and flag approaching deadlines. Tencent also plans to introduce additional team collaboration features through WorkBuddy’s upcoming Projects feature.Financial and coding advancesFinancial research represents another major application for Hy4. Tencent trained the model using thousands of specialized workspaces and tens of thousands of professional tasks. These exercises covered fund research, securities analysis, and quantitative investment strategies.In a blind evaluation involving 213 financial professionals, Hy4 achieved the highest overall score. The assessment included 579 real-world financial tasks. Tencent said approximately 70 percent of participating professionals had over five years of industry experience.Hy4 scored 3.04, narrowly ahead of DeepSeek-V4.1-Flash-1017 at 2.97. Kimi K3.1 Pro and GLM-5.3 both recorded 2.68.Software development also received attention during training. Hy4 improved its GameDev benchmark score by 19.1 percent over Hy4 Preview. Its GameLogic score increased by another 9.8 percent.The model supports development workflows involving Unreal Engine, Unity, and Godot. Tencent says these improvements extend to complex game mechanics and multiplayer implementations.Lower costs and safety measuresTencent has also worked to reduce factual errors when Hy4 retrieves information. Internal evaluations showed its hallucination rate falling from 12.95 percent to 7.2 percent. Errors linked to unsupported assumptions dropped from 23.1 percent to 5.6 percent.Cybersecurity safeguards form another part of the release. Tencent trained Hy4 on defensive security material, including vulnerability analysis and remediation. The training excludes offensive exploitation materials, according to the company.Hy4 is available through Tencent Cloud TokenHub, WorkBuddy, CodeBuddy, Yuanbao, and several other platforms. Its API also supports third-party services, including OpenRouter and OpenCode.Pricing remains unchanged from Hy4 Preview despite the reported performance improvements. Tencent charges approximately $0.90 per million input tokens and $2.70 per million output tokens. A cached input million costs approximately $0.045.Tencent also plans to release Hy4 Flash soon. The smaller offering will target applications where lower operating costs matter most.