China’s National Supercomputing Platform Unleashes Full-Power DeepSeek-R1 via API

Beijing, February 9, 2025 – In a move poised to empower developers and enterprises, the National Supercomputing Internet Platform has announced the availability of the full DeepSeek family of models, including the highly anticipated full-powered DeepSeek-R1, accessible through API calls and AI web application services.

According to an announcement on the official Supercomputing Internet WeChat account, this launch aims to meet the growing demands of the AI community by providing access to cutting-edge large language models (LLMs) via a robust and easily deployable infrastructure.

The platform now supports rapid deployment of private API services for the DeepSeek-R1-671B full-powered version. Furthermore, it offers AI Web application deployment services for smaller versions of the R1 model, specifically the 32B variant. Users can directly leverage API call interfaces and customize deployment services by modifying image files.

[Image of DeepSeek-R1 model image creation interface on the Supercomputing Internet platform]

[Image of DeepSeek-R1 model image rapid deployment private API call service interface]

The National Supercomputing Internet Platform is rapidly expanding its reach, connecting over 20 supercomputing and intelligent computing centers across 14 provinces in China. Currently, the platform boasts over 6,500 computing power products, including nearly 240 AI model services. These services encompass a wide range of models, including domestically developed open-source models like DeepSeek and Qwen, as well as international open-source models such as Llama, Stable Diffusion, and Gemma.

Specifically, the platform offers the following deployment options for DeepSeek-R1:

DeepSeek-R1 (1.5B, 8B, 70B, 671B): Supports rapid deployment of private API services.
DeepSeek-R1 (7B, 14B, 32B): Supports deployment of AI Web applications.

This launch marks a significant step in democratizing access to advanced AI capabilities, enabling developers and enterprises to integrate powerful LLMs into their applications and workflows with greater ease and efficiency. By providing a comprehensive platform with diverse model options and flexible deployment methods, the National Supercomputing Internet Platform is solidifying its position as a key enabler of AI innovation in China.

Conclusion:

The National Supercomputing Internet Platform’s launch of the full-powered DeepSeek-R1, coupled with API support and web application deployment options, represents a crucial advancement in making powerful AI models accessible to a wider audience. This initiative is expected to accelerate AI development and adoption across various industries in China, fostering innovation and driving economic growth. Future research and development efforts should focus on further optimizing the platform’s performance, expanding the range of available models, and enhancing the user experience to maximize its impact on the AI ecosystem.

References: