Anthropic is giving its Claude 3 LLMs the new ‘tool use’ feature, which allows users to build their own AI-driven solutions. Known as ‘AI agents,’ these can perform certain functions fully automated, providing a more personalized experience.
The new feature, ‘tool use,’ for the Claude 3 LLMs allows users to create and use an AI assistant for personalized applications. They contact any external API or other tool of their choice and then have the tool perform the associated action.
Examples of tool use include analyzing data to make personalized product recommendations or quickly answer customer questions, tracking an order, setting up an email assistant or providing real-time technical support.
Support for developers
Using tool use does require some specific knowledge. In any case, it requires access to APIs and, in addition, a reasonable understanding of coding.
Combining tool use with streaming functionality reduces response waiting times, allowing for more interactive interactions. In addition, “forced tool use” helps instruct developers on the Claude 3 LLM used to select the appropriate tools to deploy. With this, they can determine which tool the particular LLM should or should not use or indicate that the choice is left to Claude 3. This, in turn, allows more specific and efficient applications to be developed. Furthermore, the tool supports using images to create applications that can analyze visual data.
The new AI feature is now generally available through Anthropic’s Messages API and AI hubs Amazon Bedrock and Google Vertex AI. The cost of the tool is based on the number of tokens the respective Claude 3 LLM processes.
Also read:‘Claude 3 is better than GPT-4 and Gemini’: OpenAI has more and more competitors