Deep Agents Deploy: an open alternative to Claude Managed Agents

LangChain is launching Deep Agents Deploy in beta, an open-source, model-agnostic alternative to Claude Managed Agents for deploying AI agents in production. A single `deepagents deploy` command handles multi-tenant orchestration, sandbox setup, and exposes 30+ endpoints including MCP, A2A, Agent Protocol, human-in-the-loop, and memory endpoints. The core argument is about avoiding vendor lock-in: proprietary harnesses like Claude Managed Agents bundle memory behind closed APIs, making it costly to switch models or platforms. Deep Agents stores memory in open formats (AGENTS.md, skills files), supports every major model provider and sandbox provider, and is MIT-licensed for both Python and TypeScript. Self-hosting LangSmith Deployments keeps memory fully under the user's control.