Microsoft has announced that DeepSeek R1 is now available through Azure AI Foundry and GitHub.
DeepSeek is an open source AI model created in China that has been a topic of conversation over the past week after claims of how the model was seemingly trained for a fraction of what U.S.-based AI companies have spent training their models and how it uses less computing power to run.
Sales of Nvidia’s high performing chips were restricted in China, forcing DeepSeek to train these models on the lower-performance H800 chips. This has other companies considering the possibility that high-end chips may not be as necessary for AI development as previously thought. The R1 model competes on performance and capabilities with models from OpenAI, Meta, and Google, but with significantly lower costs.
“As part of Azure AI Foundry, DeepSeek R1 is accessible on a trusted, scalable, and enterprise-ready platform, enabling businesses to seamlessly integrate advanced AI while meeting SLAs, security, and responsible AI commitments—all backed by Microsoft’s reliability and innovation,” Asha Sharma, corporate vice president of Microsoft’s AI Platform, wrote in a blog post.
The addition to Microsoft’s Azure AI Foundry and GitHub will enable developers to experiment with DeepSeek R1 while using Microsoft’s built-in model evaluation tools to compare outputs and benchmark performance.
Microsoft also says it has put the model through red teaming and safety evaluations, including automated assessments of model behavior and security reviews to mitigate potential risks. Azure AI Content Safety provides built-in content filtering by default, with the option to opt-out, and the Safety Evaluation System allows applications to be tested before they are deployed.
“These safeguards help Azure AI Foundry provide a secure, compliant, and responsible environment for enterprises to confidently deploy AI solutions,” Sharma wrote.
