American AI start-up Anthropic has revealed its most capable large language model (LLM) yet, Sonnet 3.7, claiming it to be the first hybrid model. The new model was announced alongside Claude Code, an AI tool for agentic coding. 

“Today, we’re announcing Claude 3.7 Sonnet1, our most intelligent model to date and the first hybrid reasoning model on the market. Along with the model, we’re also introducing a command line tool for agentic coding, Claude Code.”, the company posted on its official website. 

Claude 3.7 Sonnet can function as a normal LLM and a reasoning model, hence why Anthropic calls it a hybrid model. In its standard mode, Claude 3.7 Sonnet can answer questions like a regular chatbot. In the “Extended Thinking mode”, the model is able to reason using logic which improves its output. Compared to its predecessors, Claude 3.7 Sonnet shows superior performances in coding and front-end web development in particular. 

See Related: Anthropic’s New Claude 3.5 Sonnet The Latest AI Chatbot Claiming To Be The Best

Claude 3.7 Sonnet has an output limit of 128K, similar to Claude 3.5 Sonnet. But users can now set the amount of tokens the model can use per output. This allows for higher-quality answers.

Claude 3.7 Sonnet is available to all Claude users as well as Anthropic API, Amazon Bedrock, and Google’s Vertex AI. The Extended Thinking mode is accessible on all the platforms except Free users. 

Claude Code is only available for limited research preview. It is a tool designed specifically to help programmers with coding, such as editing, debugging, linting, helping with the logic and architecture of the code, and more.

The post Introducing Claude 3.7 Sonnet: Anthropic’s Most Capable Model Yet appeared first on The Distributed.