.agent community
  • Map
  • Events
  • About
HomeMemberLlamaindex
See poster
.agent

The open community of the people building the agentic web. Open standards, open work streams, and a public map of members. Also the applicant for the proposed .agent top-level domain, pending ICANN approval. Operated by Open Agent Registry, Inc.

Discover
  • Map
  • Events
  • Team
  • Members
Mission
  • About
  • Why join
  • Brand
  • Blog
Build
  • Docs
  • Developers
  • AID spec
  • Glossary
  • Governance
  • Lists
  • GitHub
  • npm
Legal
  • Charter
  • Terms
  • Privacy
  • DNS Abuse Policy
  • Contact
  • ICANN-safe copy
© 2026 Open Agent Registry, Inc. · .agent is a proposed TLD, pending ICANN approval.
EN·v2026.04
Map·LlamaIndex
LlamaIndex

LlamaIndex

The new standard for complex document processing

See the posterShareable periodic grid→
Member since
2026
Location
San Francisco, CA
Team
11-50
Founded
2022

Links

  • www.llamaindex.ai
  • GitHub
  • LinkedIn
  • @llama_index
  • Blog
  • Docs
  • Discord
Role in the agent ecosystem

LlamaIndex is a fundamental layer in the agent stack, specifically focused on the "knowledge" and "context" components. For an agent to be useful in a corporate setting, it must access and interpret private documentation. LlamaIndex provides the tools to turn unstructured data into "Agent Skills," which are essentially specialized functions that an agent can call to answer questions or perform tasks based on a specific knowledge base.

The company is pushing the concept of "Agentic RAG," where the retrieval process itself is handled by an agent. This allows for more sophisticated behaviors, such as an agent deciding to look up a different document if the first one was insufficient. Their recent support for the Model Context Protocol (MCP) further cements their role in the ecosystem, making it easier for coding agents and other AI tools to search through documentation and data sources natively.

About

The document data interface

LlamaIndex began as GPT Index, an open-source project by Jerry Liu to help developers connect their own data to OpenAI's models. While large language models are capable of reasoning, they are inherently limited by their training cutoff and lack of access to private enterprise files. LlamaIndex solves this by providing a modular framework for Retrieval-Augmented Generation (RAG). The company has since evolved from a simple indexing tool into a complete ecosystem for building agents that can read, reason, and act on complex documents.

Solving the PDF problem with agentic OCR

One of the most significant challenges in building AI agents is document parsing. Traditional OCR tools often fail when encountering complex spatial layouts, such as nested tables in financial statements or hierarchical headers in medical records. LlamaIndex addresses this through LlamaParse, a managed service that uses "agentic OCR." Unlike static parsers, this system uses vision models to understand the structural intent of a document, converting it into a format that a language model can actually use. This technical wedge is how the company moves from an open-source framework into a paid enterprise platform.

Modular building blocks for agentic workflows

The LlamaIndex architecture is built around several core abstractions: data connectors to ingest information, indices to structure it, and query engines to retrieve it. For agentic use cases, the framework provides Workflows, an event-driven architecture that allows developers to build stateful systems. These systems can pause for human review, reflect on their own outputs, and loop through multi-step reasoning processes. This approach is distinct from linear chains; it allows for non-linear, async-first applications that scale more reliably in production environments.

Founders and market position

Founded in 2022 by Jerry Liu and Simon Suo, the company is based in San Francisco and has secured significant backing from Greylock and Norwest Venture Partners. In the competitive landscape, LlamaIndex is frequently compared to LangChain. While LangChain is often used for its broad library of agent logic and general-purpose tools, LlamaIndex has carved out a specialty in data management. It is the framework choice for developers who prioritize the "Retrieval" in RAG. This focus has led to adoption among financial institutions for due diligence automation and healthcare providers for clinical record processing. The company maintains a dual strategy: a popular open-source library that drives community adoption and a managed LlamaCloud platform designed for enterprise-grade security and scale.

Products
#01

LlamaIndex OSS

The leading open-source framework for building context-augmented LLM applications.

#02

LlamaParse

Agentic OCR and document parsing for complex enterprise files.

Open source on GitHub
Hiring

LlamaIndex is hiring.

View openings →
Similar builders
Q

Qantica

qantica.agent

B

BQX

bqx.agent

B

Browser Use

browser-use.agent

O

OmniAgent

omniagent.agent