DeepSeek-V4-Flash API launches on China’s National Supercomputing Internet

DeepSeek-V4-Flash has entered public beta, with the National Supercomputing Internet rolling out API access and model downloads on day one. Developers can access the API through the platform’s Model Services portal after logging into the National Supercomputing Internet website. The model is also available for download via the platform’s AI community, where a trusted Notebook environment enables developers to fine-tune, deploy, and adapt the model through a streamlined end-to-end workflow.