Agents

A Python framework and platform for building, running, and managing AI agents, with built-in memory, tools, and multi-agent teams.

An all-in-one, self-hostable desktop and web app for running local-first AI agents over your own documents, with RAG, tools, and multi-user support.

Microsoft's framework for building multi-agent AI applications, where several conversable agents coordinate, call tools, and hand off work to solve tasks together.

One of the earliest autonomous GPT agents, now a platform for building, deploying, and running AI agents that chain tasks toward a goal with minimal human input.

Lets AI agents control a real web browser to navigate sites and complete tasks, turning natural-language goals into automated clicks, form fills, and data extraction.

A framework for orchestrating teams of role-playing autonomous agents that collaborate on complex tasks, each with its own goal, tools, and backstory.

Open-source infrastructure for computer-use agents: sandboxes, SDKs, and benchmarks for AI that controls full macOS, Linux, and Windows desktops.

An open-source long-horizon agent harness from ByteDance that researches, writes code, and produces content using sandboxes, memory, tools, and subagents.

Google's open-source AI agent that brings Gemini into the terminal for coding, automation, and MCP-connected workflows.

A widely adopted framework for building LLM applications and agents, with composable abstractions for prompts, memory, tool calling, and retrieval across many model providers.

A library from the LangChain team for building stateful, multi-step agent workflows as graphs, with durable execution, branching, and human-in-the-loop control.

A platform for building stateful agents with long-term memory that persists across sessions, letting agents learn and improve over time (formerly MemGPT).

A deliberately minimal (~100-line) coding agent that resolves GitHub issues and assists at the command line, while scoring competitively on SWE-bench.

An agent framework from the Pydantic team that brings type-safe, structured outputs and validation to LLM applications and tool calling.

A developer-first open-source framework for building, running, and managing autonomous AI agents, with a GUI, tool integrations, and concurrent agent support.