---
id: pyroscope-io
version: "1.2.1"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# pyroscope-io — Pyroscope Python integration

License: permissive · Maintenance: active · Downloads: 11.7M/mo

## What it is and what it does

Pyroscope-io is a Python profiling agent that continuously samples your application's call stacks and streams the data to a Pyroscope server. It uses py-spy for stack sampling and a Rust backend to handle profile transmission, giving you ongoing visibility into where your application spends CPU time without requiring code instrumentation.

The package is designed for production profiling workflows where you want to understand performance characteristics over time rather than one-off profiling sessions. It has no Python runtime dependencies and works across macOS, Linux, and Unix systems on modern Python versions. Installation involves downloading platform-specific wheels, so setup is straightforward but version-specific.

Use it for:

- Identify CPU hotspots in long-running services by collecting continuous profiles over time
- Debug performance regressions by comparing flame graphs from different time periods
- Monitor production applications to catch unexpected performance degradation early
- Analyze batch job performance to optimize resource allocation and execution time
- Profile multi-threaded or async applications to understand concurrency patterns

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Pyroscope-io is a continuous profiling agent that samples Python application stack traces and sends them to a Pyroscope server for performance analysis.

Yes, if you have a Pyroscope server running and need production-grade continuous profiling. The package is actively maintained, has no external Python dependencies, and carries a permissive license. Medium install friction is typical for compiled profilers and not a barrier. No known security vulnerabilities. Install it when you're ready to move beyond ad-hoc profiling to systematic performance monitoring.

## Install

pip install pyroscope-io
uv add pyroscope-io
poetry add pyroscope-io

## Installing pyroscope-io

Before you install: Medium install friction due to platform-specific wheels (compiled bindings for multiple Python versions and architectures). Active maintenance with a recent release 18 days ago. No runtime dependencies to manage.

License in practice: Apache-2.0 permissive license allows commercial and private use with minimal restrictions; you must include a copy of the license and state significant changes.

Quickstart:

pip install pyroscope-io

import pyroscope

pyroscope.configure(
    app_name="my-app",
    server_address="http://localhost:PORT"
)

# Your application code runs with profiling enabled

Requires Python 3.10 or later. Requires a running Pyroscope server to send profiles to.

Verify before relying:

- Exact overhead and performance impact of continuous sampling on production workloads
- Whether py-spy integration requires any system-level permissions or capabilities
- Supported Pyroscope server versions and compatibility guarantees
- Default server address and port configuration for Pyroscope integration

## Package facts

- License: Apache-2.0 (permissive)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 11.7M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

python profiling agent, continuous profiling, stack sampling profiler, performance monitoring python, pyroscope integration, cpu profiling agent, application profiler, profiling, performance-monitoring, observability

[View on SkillFed](https://skillfed.io/packages/pyroscope-io) · [View on PyPI](https://pypi.org/project/pyroscope-io/)