AWS Adds DeepSeek’s R1 AI Model To Amazon Bedrock, SageMaker

'DeepSeek R1 is the latest foundation model to capture the imagination of the industry,' said AWS CEO Matt Garman.

Cloud and AI giant Amazon Web Services has enabled DeepSeek’s newest R1 foundation model to be used within its flagship AI platform Amazon Bedrock and Amazon SageMaker.

“We’ve always been focused on making it easy to get started with emerging and popular models right away, and we’re giving customers a lot of ways to test out DeepSeek AI,” said AWS CEO Matt Garman in a LinkedIn post.

AWS customers can now deploy DeepSeek’s R1 Distill Llama models on Amazon Bedrock. Bedrock customers can utilize AWS’ Custom Model Import feature, which allows users to integrate their fine-tuned models—like DeepSeek R1— seamlessly into the Bedrock environment.

“Customers can run the distilled Llama and Qwen DeepSeek models on Amazon SageMaker AI, use the distilled Llama models on Amazon Bedrock with Custom Model Import, or train DeepSeek models with SageMaker via Hugging Face,” said AWS’ CEO.

[Related: AWS CEO On President Trump, New AI Model And $8.3B Cloud Infrastructure Investment]

The $110 billion Seatle-based AWS has poured billions into AI hardware and innovation over the past several years.

For example, just this month, AWS launched its new video Luma AI model for Bedrock as well as a $8.3 billion investment in AI and cloud infrastructure in India.

“DeepSeek R1 is the latest foundation model to capture the imagination of the industry,” said Garman.

DeepSeek Shakes Up AI Market

DeepSeek—which offers similar technology to OpenAI’s ChatGPT—has taken the AI world by storm in 2025.

The Chinese AI startup released its newest AI model R1 this month, which has been hailed as a game changer due to its AI benchmark performance alongside its training price. AI training can be extremely costly due to the price of GPUs.

While OpenAI’s GPT-4 training cost was upwards of $100 million, DeepSeek said R1’s cost was less than $6 million to train.

DeepSeek’s less than $6 million price tag to build R1 sent shockwaves through the industry as most AI companies pour tens of millions into building AI models. For example, AI hardware kingpin Nvidia saw a record stock market decline on Monday of approximately $600 billion due to its stock dropping 17 percent after DeepSeek unveiled R1.

AWS Partner: Adding DeepSeek Shows ‘Amazon Will Host The Hottest AI Models’

One top executive from a solution provider that’s an AWS premium tier services partner said AWS is showing a clear message that it will host the most innovative AI models on its platforms regardless of if its homegrown models or from third parties.

“AWS’ super rapid response to the DeepSeek headlines, is basically, ‘Yeah, of course it will work on AWS. We’re not worried. And guess what, for the next AI model to grab headlines—it’ll be on Bedrock too,’” said the executive who declined to be identified. “Amazon will host the hottest AI models in the market on top of AWS. That’s the position they’re taking. We love it. … We’re all for interoperability with the latest and great models for our customers.”

Garman echoed a similar message to AWS customers and partners.

“We’ve always believed that no single model is right for every use case, and customers can expect all kinds of new options to emerge in the future,” said Garman. “That’s why Amazon Bedrock and Amazon SageMaker AI make it seamless for customers to test and evaluate the latest models as they emerge and pick the best ones based on their unique needs.”

Amazon is set to report AWS’ financial earnings results for fourth quarter 2024 on Feb. 6.

Close