// MODEL OPTIMIZATION AND PROMPT SYNTAX TERM
Tool Use
When an AI model uses external programs or APIs, like a calculator or a search engine, to help it complete a task.
TECHNICAL DEFINITION
The capability of an LLM to invoke and interact with external APIs, functions, or software tools (e.g., search engines, code interpreters, databases) to augment its knowledge or perform specific actions beyond its inherent capabilities, enhancing task execution and grounding.
BACKGROUND
Prompt engineering is the process of structuring natural language inputs to produce specified outputs from a generative artificial intelligence (GenAI) model. Context engineering is the related area of software engineering that focuses on the management of non-prompt contexts supplied to the GenAI model, such as metadata, API tools, and tokens.
READ MORE ON WIKIPEDIASYNONYMS & ALIASES
- Function calling
- API integration
- external utilities
- agentic behavior
- plugin use
USAGE NOTE
Tool use allows LLMs to overcome limitations like outdated knowledge or lack of computational precision.
DEVELOPERS
Organizations developing technology related to Tool Use.
A leading AI research and deployment company that has developed capabilities like function calling, plugins, and the Assistants API, enabling their language models to interact with external tools and services.
Known for advanced AI research, Google DeepMind develops models like Gemini which demonstrate sophisticated tool-use capabilities, integrating with search, code interpreters, and other tools to solve complex problems.
Meta AI conducts extensive research in large language models and their capabilities, including significant contributions to the concept of tool augmentation for LLMs, exemplified by their research on models like Toolformer.
Through Azure OpenAI Service, Microsoft offers features like function calling, empowering developers to integrate external tools with OpenAI models. Their research divisions also contribute significantly to the field of AI agents and tool use.
An open-source framework designed to simplify the creation of applications with large language models, with a core focus on enabling agents to use tools, connect to various data sources, and perform complex reasoning tasks.
A data framework for LLM applications, LlamaIndex focuses on connecting LLMs to external data sources and tools, facilitating the creation of powerful data-aware agents and Q&A systems.
Developers of the Claude family of LLMs, Anthropic emphasizes responsible AI development and their models support advanced prompt engineering techniques, including the effective integration and use of external tools.
Hugging Face provides a vast ecosystem of models and tools for machine learning, including 'Transformers Agents' that enable developers to build LLM-powered agents capable of using various tools for diverse tasks.
AWS offers services like Amazon Bedrock and associated agent capabilities that enable developers to build applications with foundation models, incorporating tool use to extend their functionalities and interact with AWS services and custom APIs.