If there’s one thing we’ve learned at AWS, it’s the importance of experimentation. When you’re creating something new, it’s crucial to be able to try out different technologies and quickly iterate on an idea. But experimentation can be expensive, especially for a scrappy startup team at the earliest stages of innovating. That’s one reason why we launched AWS Activate, a program focused on supporting startup founders in every step of their journey, including providing more than $6 billion in credits to help you and others like you experiment on the AWS cloud with little-to-no upfront cost.

Today, we’re taking another step to make it even easier for founders to build and iterate on their solutions using the latest technologies. We’re making AWS Activate credits redeemable for third-party models on Amazon Bedrock, our fully-managed service that offers a choice of high-performing foundation models (FMs) from leading artificial intelligence (AI) companies, like AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI, and Amazon via a single API. This means founders everywhere can now use their AWS Activate credits to experiment with these and other FMs, along with a broad set of capabilities needed to build responsible generative AI applications with security and privacy. Our goal is to make it easier for startups to evaluate what FMs are more appropriate for their use cases and find the perfect match.

Helping startups innovate with generative AI

With our full-stack generative AI offering, AWS is helping more companies around the world embrace the potential of generative AI to transform customer experiences, enhance people’s productivity, and discover new business opportunities. Expanding AWS Activate credits to Amazon Bedrock is a key way to help startups leverage generative AI in their solutions from the start.

This is one of the many benefits we have identified by working backwards from the needs of our customers, like Y Combinator. Since 2005, Y Combinator has funded and helped founders launch, build, and scale over 3,000 companies—including 60 unicorns—which currently have a combined valuation of more than $600 billion.

“With virtually every startup quickly becoming an AI startup, our partnership with AWS has never been more relevant to the companies getting into our program,” said Michael Seibel, Group Partner at Y Combinator. “AWS has been a long-standing partner and a relentless advocate for our founders, helping them with hands-on support and access to the tools they need to build the products and services people all over the world use and love.”

Since 2009, AWS has provided hundreds of millions of dollars and technical support to Y Combinator companies such as Stripe, Brex, Rappi, and many more—and in the last three years alone, AWS has provided more than $125 million in credits to Y Combinator startups. Our high-touch approach has resulted in Y Combinator companies consistently choosing AWS as their cloud provider—more than 70% of Y Combinator-funded companies run on AWS and if you look at the last two years alone, when the use of AI/ML has become more prevalent among startups, that number jumps to 80%.¹

More than 80% of Y Combinator-funded companies run on AWS (per Feb 2024 numbers for 2022-2023 batches)

For the latest Y Combinator cohort (January 2024), AWS put together an exclusive package of benefits to help startups reduce upfront costs and get access to reliable, high-performance infrastructure to build their generative AI applications on. This includes $500,000 in AWS credits that can be used for:

AWS Trainium, our purpose-built chip for training deep learning models, which offers up to a 50% cost-to-train savings over comparable Amazon Elastic Compute Cloud (Amazon EC2) instances;
AWS Inferentia, a chip designed to enable models to generate inferences more quickly and at lower cost, with up to 40% better price performance;
Reserved capacity of up to 512 NVIDIA H100 GPUs via Amazon EC2 through Capacity Blocks for Machine Learning, which dramatically increases GPU availability and ensures startups have reliable, predictable, and uninterrupted access to the GPU compute capacity required for their critical machine learning (ML) projects;
and now third-party FMs on Amazon Bedrock.

Foundation models for all

Any startup can join AWS Activate and apply for up to $100,000 in AWS credits. In addition to experimenting with Amazon Bedrock, AWS Activate credits can be used to offset costs of AWS services, including infrastructure technologies like compute, storage, databases, AI/ML, and more. Learn more and become an AWS Activate member at startups.aws.

¹Numbers as of February 2024.

AWS Activate credits now accepted for third-party models on Amazon Bedrock