# plasma-umass/scalene

Scalene: a high-performance, high-precision CPU, GPU, and memory profiler for Python with AI-powered optimization proposals

Repository: https://github.com/plasma-umass/scalene
Canonical: https://ross.abutalabs.com/products/scalene
Language: Python
License: Apache-2.0
License Family: permissive
Topics: python, profiling, performance-analysis, cpu-profiling, profiler, python-profilers, gpu-programming, scalene, profiles-memory, performance-cpu, cpu, memory-allocation, gpu, memory-consumption
Last push: 2026-08-01T19:34:24+00:00

## Health v2 (maintenance only)
Score: 92/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 95, release rhythm 83, longevity 100
- inputs: {"age_days": 2451, "days_push": 32, "days_rel": 113, "gap_med": 20, "n_releases_24m": 20}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 13495, forks 435 (observed 2026-08-28T04:11:03.414551+00:00)

## What it is
Scalene is a high-performance CPU, GPU, and memory profiler for Python that runs much faster than many alternatives while providing detailed line-level metrics. It also offers AI-powered optimization suggestions to help identify and fix performance inefficiencies.

## Use cases
- profile python script cpu usage
- find memory leaks in python code
- profile gpu usage in python
- speed up slow python program
- get ai suggestions to optimize python code
- line-level profiling of python functions

## When to choose
- you need combined CPU, GPU, and memory profiling for Python with low overhead
- you want AI-assisted optimization proposals
- you need line-level precision on where time and memory are spent

## When to avoid
- you need profiling for languages other than Python
- you require distributed-systems or production continuous profiling rather than local script profiling
- you cannot use cloud AI services for the optimization suggestions feature

## Facets
- artifact type: cli-tool
- maturity: active
- function: benchmarking, monitoring, developer-tools
- domain: developer-tools, performance
- platform: python, cli, cross-platform, windows
- tags: profiler, cpu-profiling, memory-profiling, gpu-profiling, ai-optimization, python-profiler, command-line, macos, linux

## Member repositories
- plasma-umass/scalene (main) score 92

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:03.414551+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:13:15.670256+00:00, confidence not recorded.
  - readme: https://github.com/plasma-umass/scalene (fetched 2026-08-28T04:11:03.414551+00:00, sha 895f4ccfa170)
- Data as of 2026-08-30T08:39:29.467469+00:00.