Build agents that survive production

Orchestration, memory, and eval loops for multi-step agents, past the demo and into the on-call rotation.

Production agent runtime

Individual products and alternatives (12)

CrewAI

Orchestrate the agent

CrewAI is an open-source framework for orchestrating role-playing AI agents that work together to accomplish complex tasks through collaborative multi-agent workflows.

LangGraph

Orchestrate the agent

Library for building stateful multi-actor applications with LLMs, built on top of LangChain for agent workflows.

AG2

Orchestrate the agent

Open-source multi-agent framework forked from AutoGen with improved agent collaboration and tool use.

Composio

Connect governed tools

Composio is a platform that connects AI agents to 250+ external tools and services, providing managed authentication, actions, and triggers for building production-grade AI agent integrations.

Nango

Connect governed tools

Nango is a product integrations platform that helps SaaS companies build and maintain native integrations with third-party APIs, handling OAuth, syncing, and webhook management.

Pipedream

Connect governed tools

Pipedream is a developer-first integration platform that enables building workflows that connect APIs with built-in code steps, authentication management, and event-driven triggers.

Mem0

Persist agent memory

Mem0 provides a memory layer for LLM applications that enables AI assistants and agents to remember and learn from user interactions over time, creating persistent, personalized AI experiences.

Zep

Persist agent memory

Long-term memory platform for AI assistants storing user context for personalized conversations.

Letta

Persist agent memory

Stateful AI agent framework formerly MemGPT, enabling agents with persistent memory and self-editing context for long-running tasks.

LangSmith

Evaluate the runtime

LangSmith is LangChain's platform for debugging, testing, evaluating, and monitoring LLM applications and agent workflows.

Braintrust

Evaluate the runtime

Braintrust is an AI evaluation platform for testing prompts, models, and application behavior with production-like datasets and scoring workflows.

Langfuse

Evaluate the runtime

Langfuse is an open-source platform for tracing, prompt management, evaluations, and observability across LLM applications and agent systems.