National Supercomputing Internet launches DeepSeek-V4-Flash official version API: one-click invocation, no configuration needed

Recently, DeepSeek announced that the DeepSeek-V4-Flash official version API has entered public beta testing.

Subsequently, the National Supercomputing Internet announced that the DeepSeek-V4-Flash official version model DeepSeek-V4-Flash-0731 has been officially launched, with API call services and model files provided simultaneously.

According to the introduction, the National Supercomputing Internet now opens the model's API call service to enterprises and developers. Without tedious environment configuration, users can quickly integrate and use it, providing high-performance, secure, reliable, and inclusive full-scenario AI development support for various industries.

Users can log in to the National Supercomputing Internet official website on PC, and enter the “Model Services” section through the “Services” dropdown menu on the homepage to directly access the model API call page.

In terms of model capabilities, DeepSeek-V4-Flash-0731 has undergone extensive post-training, with significant improvements in both agent capabilities and instruction-following capabilities.

In multiple benchmark tests, the DeepSeek-V4-Flash official version performs excellently, with comprehensive performance comparable to current top-tier closed-source models.

It is understood that the National Supercomputing Internet AI community currently hosts over 1,700 popular open-source large models from domestic and international sources, covering mainstream model series such as DeepSeek, GLM, Qwen, and Kimi.