DeepSeek's AI model now available on Microsoft's Azure, GitHub platforms
What's the story
Microsoft has added DeepSeek's R1 model into its Azure AI Foundry platform and GitHub. The move is likely to improve Microsoft's customer experience. This affordable AI model from the Chinese start-up has shaken US financial markets with its performance and low training costs against top models from OpenAI, Meta, and Google. The R1 model is available via a catalog on Azure AI Foundry and GitHub for easy integration into different AI apps.
Development boost
R1 model's integration accelerates AI development
Asha Sharma, Microsoft's Corporate Vice President of AI platform, highlighted the advantages of integrating models like DeepSeek R1 into Azure AI Foundry. In a blog post, she emphasized that the move would significantly speed up the pace at which developers can experiment, iterate, and incorporate AI into their workflows. The R1 model has also undergone rigorous red teaming and safety evaluations to ensure its reliability and security within the platform.
Upcoming release
Microsoft to launch compact version of R1 model
Microsoft is gearing up to launch a slimmed-down version of the R1 model for local use on Copilot+ PCs. This move suggests that we could see the R1 model being integrated into other AI-powered services from Microsoft in the future. The integration of DeepSeek's R1 model into Azure AI Foundry and GitHub was done quickly after its release as an open source model earlier this month.
Market impact
Cost-effectiveness shocks Wall Street
The R1 model has been making headlines for its cost-effectiveness, requiring fewer chips from suppliers like NVIDIA and being cheaper to train. The development has caused NVIDIA's market valuation to plummet by nearly $600 billion at one point. The decline was caused by investor fears over DeepSeek's advancements and the increasing popularity of its mobile app.
Information
Microsoft and OpenAI probe DeepSeek's training methods
Microsoft and OpenAI are also investigating whether DeepSeek used OpenAI's API to train its models. Bloomberg reported that Microsoft's security researchers had flagged large volumes of data being used through OpenAI developer accounts late last year. This usage is believed to be associated with DeepSeek.