DeepSeek R1 launches on Nvidia, AWS, GitHub, and Azure, expanding access for developers
DeepSeek R1 is now available on Nvidia, AWS, GitHub, and Azure, enhancing access for developers and businesses. This model supports advanced reasoning and is part of a growing collection of over 3,374 DeepSeek-based models on Hugging Face. AWS offers DeepSeek R1 through Bedrock, allowing easy API integration and customization. The model features 671 billion parameters and can generate up to 3,872 tokens per second, utilizing Nvidia's advanced technology for improved performance. Microsoft's Azure AI Foundry and GitHub also support DeepSeek, providing secure integration options. The model is noted for its cost efficiency, having been trained for $6 million, significantly less than similar models.