zoom_in Click to enlarge

LlamaIndex Agents

▲ ▼

language

description LlamaIndex Agents Overview

This represents the dedicated agent module within the LlamaIndex ecosystem. It focuses specifically on giving the agent the ability to use tools (like database lookups or API calls) based on the context retrieved from indexed data. It is a specialized, powerful extension of the core LlamaIndex framework, making it the go-to choice when the agent's primary job is to synthesize answers from multiple data sources via tool use.

help LlamaIndex Agents FAQ

What makes LlamaIndex Agents different from a normal LlamaIndex query engine?

A query engine retrieves or synthesizes an answer, while an agent can decide when to call tools such as a database lookup or an API. LlamaIndex documents agents as systems built from an LLM, memory, and tools.

Can LlamaIndex Agents call functions or APIs?

Yes. A LlamaIndex agent can be given tools built from ordinary Python functions, query engines, or external service wrappers. The agent selects an available tool when the user's request requires an action or lookup.

What is AgentWorkflow in LlamaIndex?

AgentWorkflow is a LlamaIndex pattern for coordinating one or more agents inside a stateful workflow. It can support multi-agent handoffs, shared state, streaming events, and human input.

Does LlamaIndex Agents work with Google Vertex AI?

LlamaIndex can be connected to hosted LLM and embedding services, including Google Cloud models through supported integrations. The exact Vertex AI setup depends on the selected model, credentials, region, and current LlamaIndex package version.

swap_horiz

Looking for LlamaIndex Agents alternatives? Compare top competitors ranked & scored

arrow_forward