Monitor your LangChain chains and agents.

A LangChain run can fan out across chains, tools and models. AgentPing rolls that up into one run record so you can see cost, status and quality at the level that matters: the agent, not the call.

Track your first AI workflow Read the Python SDK docs

Production AI in LangChain fails quietly.

LangChain makes it easy to compose chains and tools, which makes it easy to lose track of what a single run cost and whether it behaved. The provider bill is a single total; your logs show steps, not run-level cost or quality.

Wire it up in a few lines.

Initialise once, then pass the callback handler. Every tool and model call in the chain or agent is captured, and cost is priced server side from the model and token counts.

Install

install LangChain 
pip install agentping

Instrument

example · python LangChain 
import agentping
from agentping import AgentPingCallbackHandler

agentping.init(api_key="apk_…", agent="qa-chain")

# Works with LCEL chains, agents and LangGraph.
result = chain.invoke(
    {"question": "…"},
    config={"callbacks": [AgentPingCallbackHandler()], "metadata": {"customer_id": "acme-corp"}},
)

Full setup, environment variables and options are in the Python SDK docs.

What AgentPing tracks for LangChain.

  • Cost per run across every model the chain touched, priced server side.
  • Cost by agent, customer and feature, tagged at run start.
  • Run status and latency, with retries and loops made visible.
  • Missed run alerts for scheduled chains.
  • Output quality checks and rubric scoring on sampled runs.
Read the Python SDK docs Heartbeat webhook docs

What you see in the dashboard.

  • A run list for LangChain, each row showing cost, status and latency, filterable by agent, customer and feature.
  • Cost by agent, customer and feature, so you can see which accounts and features drive AI spend, not just one provider total.
  • An incident the moment a scheduled run misses its window or a cost spike breaks its baseline, delivered to Slack, email, PagerDuty or a webhook.
  • A quality trend once you turn on scoring, so a regression in your LangChain output shows up as a falling line, not a support ticket.
How do I instrument a chain?
Add the AgentPing callback handler to your LangChain run. Model calls, tool calls and the final output are captured as one run record, with cost priced on our servers from the model and token counts.
Does it cover both agents and plain chains?
Yes. Anything you run, whether an agent loop or a straight chain, becomes a run record. The unit is the run, so cost and quality roll up the same way either way.
What does it cost to monitor LangChain?
There is a free plan with no card and no time limit, enough to wire up your first agents. Paid plans start at $99/month with named limits and no metered billing, so the bill never surprises you. Annual billing is two months free.
Does AgentPing see my prompts and outputs?
Only if you choose to send them. By default AgentPing prices runs from model and token counts and tracks status and timing, which needs no prompt or output content. Payload capture for quality scoring is opt-in and configurable per agent.

Monitor your LangChain AI workflows.

Send one run from LangChain and see cost, status and latency. Then add an alert for cost spikes or missed runs.

Track your first AI workflow Read the Python SDK docs

Spend, cost attribution → Verify, quality scoring → Pulse, live monitoring →