asv

What it is and what it does

asv is a benchmarking framework designed to measure and track performance changes in Python projects across their version history. It runs a suite of benchmarks against different commits or releases, stores results, and renders them in an interactive web interface that can be hosted on any static server. The tool is built around the idea that performance regressions matter and should be visible alongside code changes.

The package depends on virtualenv to isolate benchmark environments, build and packaging for managing project builds, and pympler for memory profiling. It also uses json5, pyyaml, tabulate, and colorama for configuration, output formatting, and terminal display. This dependency chain means installation requires a working Python environment with pip and some system-level tooling, but the payoff is a complete benchmarking pipeline that integrates with version control history.

Use it for:

• Track performance regressions in a library across commits to catch slowdowns before they reach users.
• Generate a public performance dashboard for a project to demonstrate optimization work and stability.
• Compare benchmark results across Python versions to identify version-specific performance issues.
• Monitor memory usage trends in a package over time using profiling integration.
• Validate that optimization changes actually improve performance before merging to main.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

asv benchmarks Python packages over their lifetime, tracking performance changes across versions and displaying results in an interactive web interface.

Yes, if you maintain a Python package and want systematic performance tracking. The active maintenance, permissive license, and lack of known vulnerabilities make it safe to adopt. Medium install friction is reasonable for the functionality gained. Not necessary for one-off benchmarking or small projects where ad-hoc performance testing suffices.

Install

asv on PyPI

pip install asv

uv add asv

poetry add asv

Installing asv

Before you install

Medium install friction due to 11 runtime dependencies including virtualenv, build, and pympler. Actively maintained with a recent release (48 days old) and steady repository activity; supports modern Python versions (3.9+).

License in practice

BSD-3-Clause permissive license allows commercial and private use with minimal restrictions; attribution required.

Quickstart

pip install asv

# In your project root, create asv.conf.json with benchmark suite config
# Then run:
asv run

# Generate and serve results:
asv publish
asv preview

Requires Python 3.9+; virtualenv must be available; project must be in a git repository for historical benchmarking.

Verify before relying

• Whether the web frontend requires specific server configuration beyond static hosting
• Performance overhead of benchmarking itself on large test suites
• Compatibility with CI/CD platforms beyond what the fact sheet indicates

Package facts

Evidence: asv-0.6.6-cp314-cp314t-macosx_10_15_x86_64.whl; asv-0.6.6-cp314-cp314t-macosx_11_0_arm64.whl; asv-0.6.6-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl; asv-0.6.6-cp314-cp314t-musllinux_1_2_x86_64.whl; asv-0.6.6-cp314-cp314t-win_amd64.whl; asv-0.6.6-cp36-abi3-macosx_10_9_x86_64.whl; asv-0.6.6-cp36-abi3-macosx_11_0_arm64.whl; asv-0.6.6-cp36-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl; asv-0.6.6-cp36-abi3-musllinux_1_2_x86_64.whl; asv-0.6.6-cp36-abi3-win_amd64.whl

Tags

More Testing packages