Amazon Bedrock Expands AI Horizons with 100+ New Models

At this year’s AWS re:Invent, Amazon Web Services unveiled a major expansion to its Amazon Bedrock platform, introducing over 100 new AI models alongside tools designed to make generative AI more accessible and efficient for businesses. The updates are part of AWS’s broader push to simplify how companies integrate AI into their workflows, from creative content production to complex data processing.

“Amazon Bedrock is tackling the toughest challenges developers face today,” said Dr Swami Sivasubramanian, AWS’s Vice President of AI and Data. “These new capabilities help customers unlock the full potential of generative AI.”

An Expanded Model Marketplace

Central to the announcement is the Amazon Bedrock Marketplace, a new hub offering businesses a diverse selection of models to suit specific needs. From Luma AI’s Ray 2, which enables realistic video creation, to poolside’s malibu and point models, which streamline software engineering tasks, the marketplace delivers options for nearly every industry.

Stability AI’s Stable Diffusion 3.5 Large, also part of the new lineup, focuses on generating high-quality images from text descriptions. Whether a company is creating marketing assets or visual effects, the model offers flexibility and efficiency.

Customers can easily browse and deploy these models through Bedrock’s unified interface, streamlining what has traditionally been a resource-intensive process.

Smarter AI at Lower Costs

Two features, Prompt Caching and Intelligent Prompt Routing, aim to make AI usage more cost-effective. Prompt Caching eliminates redundant computations, reducing latency by up to 85% and cutting costs by as much as 90%. Meanwhile, Intelligent Prompt Routing dynamically directs queries to the most appropriate model based on cost and complexity.

Argo Labs, a developer of AI-driven voice assistants, has already embraced these tools to optimise customer interactions. “By routing simpler queries to smaller models and reserving complex questions for more advanced ones, we’ve saved time and money without sacrificing quality,” an Argo Labs spokesperson shared.

Transforming Data into Insights

A standout feature of Amazon Bedrock’s update is its ability to process unstructured data — everything from PDFs to audio recordings—into usable formats for analysis and AI applications. For instance, Amazon Bedrock Data Automation enables businesses to extract and organise data from documents, videos, and more, significantly reducing manual workloads.

BMW Group has used Amazon Bedrock’s GraphRAG capabilities to power its internal AI assistant, helping employees navigate vast stores of data with greater ease and accuracy. Similarly, digital asset firm Tenovos has leveraged Bedrock’s tools to boost content reuse by over 50%, saving millions in marketing costs.

A Growing Ecosystem

The adoption of Amazon Bedrock is accelerating. Tens of thousands of customers, from Adobe to Zendesk, are using the platform to drive innovation. With a 4.7x increase in users over the past year, AWS’s investment in generative AI appears to be paying off, positioning Amazon Bedrock as a vital tool in a fast-changing technological landscape.

Read next: AWS Trainium2 Unlocks New AI Performance Levels