LlamaParse MCP Update: New Extract & Index v2 Tools for Agents
Blog post from LllamaIndex
Recent updates to the LlamaParse MCP aim to enhance document processing by introducing new tools and a restructured server architecture, enabling more efficient workflows for developers and agents. The platform now features Extract service integrations, allowing agents to generate JSON schemas and extraction rules for structured data extraction, addressing previous limitations in document access and underspecified extraction processes. Furthermore, the introduction of Index v2 offers a managed knowledge base service with advanced retrieval capabilities, supporting a broader range of content types and allowing agents to interact directly with source documents. The server architecture has been reorganized into product-specific MCP servers, each focused on a particular document-processing task, promoting modularity and parallelism in agentic workflows. These improvements reflect a shift towards a composable approach to document processing, providing users with the tools to build robust pipelines for tasks such as classification, extraction, and retrieval over diverse content types.
No tracked trend matches for this post yet.