How to beginner · 3 min read

AgentOps pricing

Quick answer

AgentOps offers a usage-based pricing model with a free tier that covers basic AI agent observability features. You pay based on the volume of tracked API calls and sessions, with detailed pricing available on the official AgentOps website. Use the agentops Python package to integrate and monitor your AI agents efficiently.

PREREQUISITES

Python 3.8+
AgentOps API key
pip install agentops

Setup

Install the agentops Python package and set your API key as an environment variable for secure authentication.

bash

pip install agentops

# Set environment variable in your shell
export AGENTOPS_API_KEY="your_agentops_api_key"

output

Collecting agentops
  Downloading agentops-1.0.0-py3-none-any.whl (10 kB)
Installing collected packages: agentops
Successfully installed agentops-1.0.0

Step by step

Use the agentops Python SDK to initialize the client and start tracking AI agent sessions. This example shows a simple usage pattern with automatic tracking of OpenAI API calls.

python

import os
import agentops

# Initialize AgentOps with API key from environment
agentops.init(api_key=os.environ["AGENTOPS_API_KEY"])

# Start a manual session
session = agentops.start_session(tags=["example-agent"])

# Simulate AI agent work
print("Running AI agent...")

# End the session with a status
agentops.end_session("Success")

print("AgentOps session tracked successfully.")

output

Running AI agent...
AgentOps session tracked successfully.

Common variations

You can use AgentOps with automatic instrumentation of OpenAI SDK calls or manually track sessions and events. The SDK supports async usage and integrates with other AI frameworks for observability.

python

import asyncio
import agentops

async def async_agent_work():
    agentops.init(api_key=os.environ["AGENTOPS_API_KEY"])
    session = agentops.start_session(tags=["async-agent"])
    print("Running async AI agent...")
    await asyncio.sleep(1)  # Simulate async work
    agentops.end_session("Completed")
    print("Async session tracked.")

asyncio.run(async_agent_work())

output

Running async AI agent...
Async session tracked.

Troubleshooting

If you see authentication errors, verify your AGENTOPS_API_KEY environment variable is set correctly.
For missing telemetry data, ensure agentops.init() is called before any AI agent calls.
Check network connectivity if sessions fail to report.

✅

Key Takeaways

AgentOps pricing is usage-based with a free tier for basic observability.
Set your API key via environment variable and install the agentops package to start.
Use agentops.init() early in your app to enable automatic tracking.
Supports both synchronous and asynchronous Python usage patterns.
Check official AgentOps docs for detailed pricing and limits.

Verified 2026-04

Verify ↗