infiniflow/ragflow

Features

• Autonomous Agents - Integrates large language models with custom knowledge bases and external tools to execute complex, multi-step autonomous workflows.
• Chat Assistants - Exposes API endpoints for creating and managing conversational AI assistant instances.
• Retrieval-Augmented Generation Platforms - Delivers a comprehensive environment for building, managing, and deploying knowledge-based AI applications with advanced document parsing and retrieval capabilities.
• Grounded Answer Generation - Generates responses with traceable citations and visual chunking to reduce hallucinations and facilitate human verification of content.
• RAG Pipelines - Coordinates multi-stage recall, re-ranking, and citation-based generation to produce grounded, verifiable responses from indexed datasets.
• Conversational and Voice Interaction - Enables the deployment of conversational agents that leverage indexed knowledge bases to provide context-aware, human-like interactions.
• Agent Management APIs - Handles the lifecycle of autonomous agents through dedicated API endpoints for listing, managing, and interacting with system entities.
• Document Knowledge Extraction - Processes unstructured data using deep document understanding to extract structured knowledge for high-quality information retrieval.
• Knowledge Dataset Managers - Organizes knowledge by uploading, parsing, and indexing documents into structured datasets for retrieval-augmented generation.
• Semantic Search Engines - Executes semantic searches across indexed datasets to retrieve relevant information and document snippets for answering complex queries.
• Agentic Tool-Use Frameworks - Empowers agents to perform multi-step reasoning tasks by bridging internal memory with external tools and knowledge sources.
• OpenAI-Compatible APIs - Standardizes HTTP endpoints for chat completions to ensure compatibility with common AI model integration interfaces.
• Chat Assistant Management APIs - Provides API endpoints for listing, filtering, and retrieving metadata about configured chat assistants.
• Knowledge Graph Construction - Automates the construction of knowledge graph structures from datasets via dedicated API endpoints.
• Document Chunking Strategies - Segments source documents using explainable templates to optimize retrieval accuracy during knowledge base indexing.
• Chat and API Access - Maintains multi-turn dialogue context and streams model outputs via interactive chat interfaces and programmatic endpoints.
• Graph-Based Knowledge Indexers - Builds hierarchical summaries and multi-layered knowledge graphs to enhance reasoning over extensive document collections.
• Local LLM Configurations - Configures local inference engines and external model providers through a unified interface for seamless deployment.
• Knowledge Graph APIs - Facilitates the retrieval and management of dataset-specific knowledge graph structures through dedicated API endpoints.
• Document Parsing Services - Offers programmatic methods to asynchronously parse and extract content from various document types for further processing.
• AI-Powered Extraction Engines - Employs machine learning to accurately isolate structured data, tables, and text from complex document layouts for retrieval.
• Source-Based Execution Environments - Supports running the platform directly from source code to facilitate real-time debugging and local development testing.
• Configuration Management - Adjusts environment parameters to manage application behavior, resource allocation, and backend operations.
• System Service Configurations - Utilizes YAML templates to define core platform service dependencies, including database connections, object storage, and authentication providers.
• Orchestration and Multi-Agent Systems - Deploys autonomous agents that leverage memory, tools, and knowledge to complete complex, multi-step reasoning workflows.
• Development Frameworks and Tools - Open-source RAG (Retrieval-Augmented Generation) workflow platform.
• RAG and Data Pipelines - RAG engine fusing retrieval with agentic capabilities.
• Retrieval Augmented Generation - Engine focused on deep document understanding for retrieval.
• Databases and RAG - RAG engine based on deep document understanding.
• RAG Frameworks - Deep document parsing and RAG engine with multi-path retrieval.
• Knowledge Graph Orchestrators - Combines relational data structures with vector-based search to improve context-aware response generation.
• Document Parsing Pipelines - Parses diverse file formats into structured text chunks using advanced layout analysis and OCR for downstream analysis.
• OpenAI-Compatible Inference Servers - Provides an API layer compatible with standard interfaces to ensure interoperability with existing large language model ecosystems.
• Dataset Management APIs - Modifies dataset configurations programmatically through dedicated administrative endpoints.
• RESTful APIs - Exposes core platform functionality through a comprehensive suite of HTTP interfaces for external application integration.
• Knowledge Graph Deletion - Purges relational knowledge graph structures associated with specific datasets via targeted API calls.
• Semantic Parsing Tools - Interprets complex documents to produce traceable, cited answers that reduce hallucinations in retrieval tasks.
• Document Management APIs - Updates document configurations and storage settings programmatically through a structured management interface.
• Document Retrieval APIs - Queries and lists documents within datasets using flexible filtering, pagination, and sorting parameters.
• Document Deletion APIs - Deletes stored documents programmatically to maintain clean and updated knowledge repositories.
• Document Parsing Controls - Terminates active document parsing tasks through specific API endpoints to manage resource utilization.
• Automated Document Ingestion - Uploads and transforms various file types into structured knowledge base entries through automated ingestion routines.
• Python SDKs - Offers a native library for Python developers to interact with platform services and manage retrieval workflows.
• RAG Pipeline Optimizers - Tunes batch processing, OCR engines, and parsing services to minimize latency in retrieval-augmented generation pipelines.
• Chat Management APIs - Manages chat assistant lifecycles, including creation, updates, and deletion, via dedicated API endpoints.
• Dataset Management - Removes datasets from the system by their unique identifiers using simple administrative commands.

Star history