Skip to main content
Aegra is an open-source, self-hosted Agent Protocol server that lets you run AI agents on your own infrastructure. It’s a drop-in replacement for LangSmith Deployments — same SDK, same APIs, no vendor lock-in. Read why teams choose Aegra over managed platforms, or jump straight to the migration guide if you’re switching.

Before you begin

Using an AI coding tool? Copy the prompt below to add the Aegra skill and MCP server for better results when building with Aegra.

What Aegra does

Aegra gives you a production-ready server for running LangGraph agents with full persistence, streaming, and authentication. You bring your own PostgreSQL database and keep complete control over your data.

Agent Protocol

Works with Agent Chat UI, LangGraph Studio, and CopilotKit out of the box.

Self-hosted

Run on your own infrastructure with your own PostgreSQL database. No external dependencies.

Same SDK

Use the LangGraph SDK you already know. Switch from LangSmith Deployments with zero code changes.

Open source

Apache 2.0 licensed. Inspect, modify, and contribute to the codebase.

How it compares

LangSmith DeploymentsAegra
Deploy agentsLocal dev only (Free), paid cloud (Plus)Free, unlimited
Custom authNot available (Free), available (Plus)Python handlers (JWT/OAuth/Firebase)
Self-hostedEnterprise only (license key required)Always (Apache 2.0)
Own databaseManaged only (Free/Plus), bring your own (Enterprise)Bring your own Postgres
TracingLangSmith onlyAny OTLP backend (Langfuse, Phoenix, etc.)
Data residencyLangChain cloud (Free/Plus), your infrastructure (Enterprise)Your infrastructure
SDKLangGraph SDKSame LangGraph SDK
Based on LangChain pricing as of February 2026. An enterprise tier with self-hosting is also available at custom pricing. See the full feature support matrix for what Aegra supports today.

Key features

  • Human-in-the-loop — Approval gates and user intervention points
  • Streaming — Real-time responses with network resilience
  • Persistent state — PostgreSQL checkpoints via LangGraph
  • Configurable auth — JWT, OAuth, Firebase, or none
  • Observability — Fan-out tracing via OpenTelemetry to multiple backends
  • Semantic store — Vector embeddings with pgvector
  • Custom routes — Add your own FastAPI endpoints alongside the Agent Protocol API

Next steps