LangGraph Tracing & Node-Level Monitoring
Auto-instrument every LangGraph graph execution, node by node, edge by edge. Get per-node span trees, cost attribution, and full visibility into which path your agent took and why.
Free tier · No credit card · 2-minute setup
What you get with Fluiq for LangGraph
Node-level span tree
Every graph node execution is a separate trace span, see which nodes ran, in what order, how long each took, and what tokens they consumed.
Per-node cost breakdown
Token and USD cost attributed per graph node. Know exactly which node in your agent graph is expensive, and track cost across long-running multi-turn graphs.
Branch & state tracing
Conditional edges and state transitions are recorded in the trace, making it easy to debug why a graph took a particular execution path.
Add Fluiq to your LangGraph app in 2 lines
import fluiq
fluiq.instrument(api_key="fl_...") # patches LangGraph automatically
from langgraph.graph import StateGraph, END
from typing import TypedDict
class AgentState(TypedDict):
messages: list
builder = StateGraph(AgentState)
builder.add_node("planner", planner_node)
builder.add_node("executor", executor_node)
builder.add_conditional_edges("planner", route_fn, {"execute": "executor", "end": END})
builder.set_entry_point("planner")
graph = builder.compile()
# Each node is a traced span, no code changes needed:
result = graph.invoke({"messages": [("user", "Analyse sales data")]})What Fluiq instruments in LangGraph
Every call to these methods is automatically traced, no decorators, no wrappers, no manual spans.
CompiledGraph.invoke()CompiledGraph.ainvoke()CompiledGraph.stream()CompiledGraph.astream()Node function calls (any add_node target)Conditional edge routing functionsMemorySaver checkpointer operationsSubgraph invocationsStart tracing LangGraph in 2 minutes
Free tier. No credit card. Full traces, security scanning, and evals on your first LangGraph call.
50,000 free traces / month · 1,000 evals / month · 14-day retention