Agents tagged agent-framework

Most popular open-source framework for building LLM applications and agents in Python and JavaScript. LangChain is the most widely adopted open-source framework for building LLM applications and agents, available in Python and JavaScript/TypeScript.

Open-source visual builder for LangChain that lets developers create LLM workflows through a drag-and-drop interface. Langflow is an open-source visual builder for LangChain that lets developers create LLM workflows through a drag-and-drop interface.

Graph-based framework for building stateful, multi-step agent workflows with explicit control flow from LangChain. LangGraph is a graph-based framework for building stateful, multi-step agent workflows with explicit control flow, built by the LangChain team.

Stateful agent framework with focus on memory, evolving from MemGPT research, building agents that learn and remember over time. Letta (formerly MemGPT) is a stateful agent framework with focus on memory and learning, evolving from MemGPT research at UC Berkeley.

Meta's open-source framework for building AI applications with Llama models including agents, RAG, and safety. Llama Stack is Meta's open-source framework for building AI applications with Llama models, providing standardized APIs for inference, agents, RAG, safety, and tools.

Data framework for building LLM applications with strong focus on RAG, document indexing, and structured data integration.

Enterprise agent platform for building autonomous AI agents with focus on production deployment and governance. Lyzr is an enterprise agent platform for building autonomous AI agents with focus on production deployment, on-prem options, and governance for enterprise use cases.

Lightweight Python AI framework from PrefectHQ for building structured LLM applications with simple ergonomic APIs. Marvin is a lightweight Python AI framework from PrefectHQ for building structured LLM applications with simple ergonomic APIs.

TypeScript agent framework for building production AI applications with workflows, RAG, and multi-agent capabilities. Mastra is a TypeScript agent framework for building production AI applications with workflows, RAG, multi-agent orchestration, and MCP support.

Memory layer for AI agents (formerly Embedchain) providing personalized context across conversations and applications. Mem0 (formerly Embedchain) is a memory layer for AI agents providing personalized context across conversations and applications.

Open-source multi-agent framework simulating a software company with AI agents playing different role-based positions.

Microsoft's agentic framework combining Semantic Kernel and AutoGen for building production multi-agent systems. Microsoft Agent Framework is Microsoft's unified agentic framework combining the strengths of Semantic Kernel and AutoGen for building production multi-agent systems.

Microsoft's framework for building LLM applications with visual flow design, evaluation, and Azure AI deployment. PromptFlow is Microsoft's framework for building LLM applications with visual flow design, prompt engineering tooling, evaluation, and deployment to Azure AI.

Educational tiny LLM framework demonstrating LLM application patterns with minimal abstraction overhead. MiniChain is an educational tiny LLM framework from Sasha Rush (Cornell) demonstrating LLM application patterns with minimal abstraction overhead.

Pythonic LLM library treating LLM calls as Python functions with strong type-safety, validation, and tools support. Mirascope is a Pythonic LLM library treating LLM calls as Python functions with strong type-safety (Pydantic), validation, tools support, and structured outputs.

Workflow automation platform with AI features and AI agent capabilities for connecting LLMs with 1000+ integrations. n8n is an open-source workflow automation platform with AI features and AI agent capabilities for connecting LLMs with 1000+ integrations.