en.Wedoany.com Reported - On July 3, Tencent Cloud announced the launch plan for the official version of the DeepSeek-V4 "Original Factory Direct" model. The model is scheduled to officially launch on Tencent Cloud's large model service platform TokenHub and the intelligent agent development platform in mid-July, with a peak-valley pricing mechanism to be introduced upon the official release.
After the DeepSeek-V4 "Original Factory Direct" model enters the Tencent Cloud platform, enterprise customers can access model capabilities through TokenHub and the intelligent agent development platform for scenarios such as intelligent Q&A, content generation, code assistance, knowledge base applications, business process automation, and intelligent agent development. TokenHub serves as the model invocation entry point and unified service platform, while the intelligent agent development platform is more aligned with enterprise application construction. For enterprises, factors such as stable model invocation, ease of integration with existing systems, clear billing, and controllable invocation peaks will all affect the speed at which large models are truly deployed into business systems.
"Original Factory Direct" means that Tencent Cloud will provide platform-based access services centered around DeepSeek's official model capabilities. When using large models, enterprise customers not only focus on the model's performance but also on interface compatibility, invocation stability, account permissions, resource scheduling, log monitoring, and cost management. By integrating the official DeepSeek-V4 version into TokenHub and the intelligent agent development platform, Tencent Cloud enables enterprises to complete model invocation, application development, and intelligent agent configuration within the cloud platform, reducing the complexity of separately interfacing with model vendors, developing interfaces, and managing operations and maintenance.
The peak-valley pricing mechanism is another key focus of this adjustment. After the official version launches, Tencent Cloud will adjust the model pricing strategy in sync with the original manufacturer, with prices differing between peak and off-peak periods. During peak hours, resource demand is more concentrated, requiring stronger computing power scheduling and service guarantees from the platform; during off-peak hours, prices are relatively lower, which is beneficial for enterprises to schedule batch tasks, offline processing, and non-real-time tasks during low-demand periods. For AI application operators, peak-valley pricing makes model invocation cost management more important, and enterprises need to adjust their invocation strategies based on business priorities, response timeliness, and budget arrangements.
Currently disclosed information focuses on the launch plan and billing adjustments for the official DeepSeek-V4 "Original Factory Direct" model, with no details yet on the complete feature list, performance indicators, enterprise customer cases, or invocation scale after the official release. What is confirmed is that Tencent Cloud plans to launch the official DeepSeek-V4 "Original Factory Direct" model on TokenHub and the intelligent agent development platform in mid-July, and will introduce the peak-valley pricing mechanism after the official release.
