LangGraph

LangGraph is LangChain's state machine layer — it uses LangChain's runnable interface, tools, and model connectors as its graph primitives.

→ Stateful, cyclical agent graphs built on LangChain's full ecosystem — every LangChain tool is a potential graph node.

LangGraph sends traces to LangSmith automatically when LANGCHAIN_TRACING_V2 is set — every node execution becomes a separate trace span.

→ Step-by-step graph execution visibility: see which nodes ran, in what order, with what inputs, outputs, and token cost.

LangGraph nodes call LiteLLM-proxied endpoints, routing each node's LLM calls to any provider without changing graph code.

→ Model-agnostic LangGraph agents — route different nodes to Claude, GPT-4o, or local models via one LiteLLM config.

LangGraph integrates with Langfuse via its callback system or OpenTelemetry, capturing every node execution as a nested trace span.

→ Full execution traces of complex agent graphs — cost per node, latency per step, and LLM call details in one view.

LangGraph routes code-execution tool calls to E2B sandboxes, giving code-generation nodes a safe isolated environment to run output.

→ Agentic code generation with sandboxed execution built into the graph — iterate on code safely without leaving the agent loop.

LangGraph retrieval nodes query pgvector via psycopg2 or SQLAlchemy, keeping vector search within the existing Postgres database.

→ Enterprise RAG with no separate vector DB — LangGraph agents retrieve from the same Postgres instance the rest of the app uses.

Maxim AI instruments LangGraph runs via trace hooks, capturing node inputs, outputs, and latency.

→ Evaluate and debug LangGraph agent workflows with structured trace replay in Maxim.

ElevenLabs TTS is invoked as a LangGraph tool node, converting agent text output to speech.

→ Add natural voice output to LangGraph agent pipelines without a separate orchestration layer.

Deepgram STT API is wrapped as a LangGraph tool node, transcribing audio at agent decision points.

→ Process voice input inline in a LangGraph agent — no separate audio pipeline required.

AssemblyAI transcription is called from LangGraph tool nodes to process audio mid-workflow.

→ Build voice-driven LangGraph agents that transcribe and reason over spoken content.

Cartesia low-latency TTS is invoked as a LangGraph tool node in real-time agent pipelines.

→ Sub-second voice output in LangGraph workflows — essential for conversational agent UX.

Vapi calls LangGraph-hosted agents as the AI backbone for voice call handling and routing logic.

→ Build voice applications where complex agentic reasoning powers the full call flow.

Retell AI connects LangGraph agents as the conversational AI engine for voice call workflows.

→ Replace static call scripts with dynamic LangGraph agent reasoning in production voice calls.

Browser Use is wrapped as a LangGraph tool node, executing web automation steps within agent workflows.

→ Build agents that browse and act on the web as structured steps in a LangGraph workflow.

Zep ships a first-party LangGraph integration that loads temporal-graph memory into each node's state and writes new facts back.

→ Add temporal, graph-structured memory to LangGraph agents with a few lines of config.