IT Brief Australia - Technology news for CIOs & IT decision-makers
Story image
Anthropic's Claude 3 AI models now available on Amazon Bedrock
Tue, 5th Mar 2024

It has been announced that the artificial intelligence safety and research company, Anthropic's Claude 3 models will be made available on Amazon Bedrock. The addition of the Claude 3 family to Amazon Bedrock is expected to enhance the capacity of users of all sizes to rapidly construct and deploy generative artificial intelligence (AI) applications across their organisations.

Amazon Bedrock currently offers customers secure and easy access to the widest selection of high-performing, fully managed large language models (LLMs) and other foundation models (FMs). It also provides the necessary capabilities for customers to safely customise these models, including best-in-class retrieval augmented generation, guardrails, model evaluation, and AI-powered agents.

The Claude 3 model range, consisting of Claude 3 Opus, Claude 3 Sonnet, and Claude 3 Haiku, will be accessible to users on Amazon Bedrock, with Sonnet available immediately, while the Opus and Haiku models will become available over the coming weeks. The introduction of Anthropic's models further expands Amazon Bedrock's range of generative AI models, boosting its capacity to drive innovation and reinvent end-user experiences.

Anthropic data has shown that Claude 3 Opus, the most intelligent model in the family, has established a new benchmark by outclassing other models, including OpenAI’s GPT-4, in reasoning, maths, and coding disciplines. The Sonnet model claims the mantle for the best blend of skills and speed, boasting double the speed of its predecessor models. The Haiku model, designed to be the market's fastest compact model, provides near-instant responsiveness and is considered one of the most cost-effective options in its intelligence category.

Anthropic also reports that Claude 3 models demonstrate significant improvements over other available models in the area of coding and understanding non-English languages including Spanish, Japanese, and French. In addition, they show lower levels of hallucinations than most popular models from other providers, and higher levels of accuracy than earlier models. These models also have advanced vision capabilities, allowing them to process and analyse image data from varied formats, providing insights from charts, graphs, technical diagrams, photos, and other visual assets.

Many global organisations of varying sizes across a range of industries are already utilising Amazon Bedrock’s generative AI applications in conjunction with Anthropic's Claude technology. This new tech is currently being used to provide automation, to create new user-facing applications, and to accelerate research and development in a host of sectors.

The collaboration between AWS and Anthropic continues to expand, with AWS becoming Anthropics' primary cloud provider for mission-critical workloads such as safety research and foundation model development. More than 10,000 global organisations are already using Amazon Bedrock to build their generative AI applications, with many more likely to adopt the new Claude 3 models.

Today’s announcement underscores AWS's continued investment in generative AI and is another step in expanding their unique offering in infrastructure, models, and user-facing applications, to enable customers to utilise generative AI in a more efficient, extensive, and integrated way across their organisations.