Skip to main content

timeit for honest micro-benchmarks

timeit helps you measure small pieces of Python code more honestly than casual timing with time.time(). It is useful for quick performance checks and micro-benchmarks.

Why it is useful​

import timeit

duration = timeit.timeit("sum(range(100))", number=10000)
print(duration)

timeit runs the code many times and is designed to reduce noise from one-off measurements.

When to use it​

  • comparing two small implementations
  • checking whether a change really helps
  • exploring performance interactively

When not to overuse it​

Micro-benchmarks can mislead if they are too far from real program behavior. Use timeit for focused questions, not as the whole performance story.

Rules of thumb​

  • Use timeit for small timing comparisons.
  • Benchmark the actual code shapes you care about.
  • Treat micro-benchmarks as clues, not final truth.