How much memory do Python threads use? Apparently not too much.

cost_of_threads.py

# Requires psutil package
import os
import threading
import time

import psutil


should_run = True


def main():
    global should_run

    report_process_mem()
    time.sleep(0.001)

    start = report_process_mem()
    thread_count = 10
    threads = []
    for _ in range(0, thread_count):
        t = threading.Thread(target=thread_runner, daemon=True)
        t.start()
        threads.append(t)

    end = report_process_mem()
    mb = (end - start)
    print(f'Started {thread_count:,} threads, consuming {mb:.01f} MB of memory total.', flush=True)

    input("Enter to exit:")
    should_run = False
    [t.join() for t in threads]


def thread_runner():
    print(f"Started thread {threading.current_thread().native_id}...", flush=True)
    while should_run:
        time.sleep(.1)

    print(f"Exited thread {threading.current_thread().native_id}.", flush=True)


def report_process_mem() -> int:
    process = psutil.Process(os.getpid())
    mb = process.memory_info().rss / 1024 / 1024
    print(f"Total memory used: {mb:,.2f} MB.")

    return mb


if __name__ == '__main__':
    main()

Output on macOS Monterey

10 threads, consuming 0.2 MB of memory total.