Tencent's Hy3 preview gives developers downloadable language-model weights and control over how much reasoning the model performs. Tencent's announcement presents it as a model for reasoning, coding and tool-using workflows.
The official model card documents direct-response and higher-reasoning settings, allowing an application to select the mode for its task. Downloadability also comes with substantial infrastructure requirements: the deployment guidance recommends eight H20-3e GPUs or GPUs with greater memory. This is a configurable model for hosted or substantial self-managed deployments, rather than an implication that local deployment is lightweight.
Terms in this piece · Glossary
mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
tool use — A model's ability to call external functions — run code, search the web, edit files — instead of only generating text.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
MCP — The Model Context Protocol — an open standard that lets any AI assistant plug into any tool or data source without custom integration code.
articleTencent Announces Global Rollout Of Scenario Based Ai Capabilities To Accelerate Industrial Efficiency
articleTencent Announces Global Launch Of Hunyuan 3d Engine To Empower Creators With Advanced Creation Tools
articleTencent Announces Global Launch Of Hunyuan 3d Engine To Empower Creators With Advanced Creation Tools
articleTencent Launches Qclaw Globally Lowering Barriers To Ai Agent Deployment
articleTencent Unveils Hy3 Preview Model Enhances Agent Capabilities And Real World UsabilityTencent
articleTencent Hunyuan Officially Releases Hy3 Advancing Agent Capabilities And Deeper Product IntegrationTencent