Enterprise CDI extension for LangChain4j — inject AI services directly into your Jakarta EE and MicroProfile applications.
Full documentation is available at langchain4j.github.io/langchain4j-cdi.
- AI Service Injection — declare AI services as CDI beans using
@RegisterAIService - Agentic Topologies — 11 per-topology annotations (
@RegisterSimpleAgent,@RegisterSequenceAgent,@RegisterLoopAgent, etc.) for multi-agent workflows - MCP Server — expose CDI beans as a Model Context Protocol server
- Configuration via Properties — configure LLM components through MicroProfile Config or a custom SPI
- Fault Tolerance — resilience with
@Retry,@Timeout,@CircuitBreaker,@Fallback - Telemetry — OpenTelemetry-based observability for AI operations
- Expression Language — resolve
${...}(MicroProfile Config) and#{...}(Jakarta EL) expressions in annotations - Guardrails — input and output validation for AI service interactions
The MCP server is dual-era: on the same /mcp endpoint it speaks both the legacy MCP 2025-03-26 protocol (session handshake) and the stateless MCP 2026-07-28 revision, so existing and modern clients connect side by side. See langchain4j-cdi-mcp/README.md for the full guide, including protocol diagrams and conformance results.
| Runtime | Extension Type |
|---|---|
| Quarkus | Build-compatible |
| Helidon | Both |
| WildFly | Portable |
| Payara | Portable |
| GlassFish | Portable |
| Liberty | Portable |
1. Add the dependency:
<!-- For portable extension (WildFly, Payara, GlassFish, Liberty) -->
<dependency>
<groupId>dev.langchain4j.cdi</groupId>
<artifactId>langchain4j-cdi-portable-ext</artifactId>
<version>${langchain4j-cdi.version}</version>
</dependency>
<!-- For build-compatible extension (Quarkus, Helidon) -->
<dependency>
<groupId>dev.langchain4j.cdi</groupId>
<artifactId>langchain4j-cdi-build-compatible-ext</artifactId>
<version>${langchain4j-cdi.version}</version>
</dependency>2. Define an AI service:
@RegisterAIService
public interface AssistantService {
@SystemMessage("You are a helpful assistant.")
String chat(String userMessage);
}3. Inject and use:
@Path("/assistant")
public class AssistantResource {
@Inject
AssistantService assistant;
@GET
@Path("/chat")
public String chat(@QueryParam("message") String message) {
return assistant.chat(message);
}
}See the documentation for configuration, agents, tools, RAG, guardrails, and more.
Example applications are in the examples/ directory, based on a car rental booking scenario. The Quarkus example is recommended for the fastest development experience.
See CONTRIBUTING.md for contribution guidelines.
Apache License 2.0 — see LICENSE file.