Test

Status: PASS (verified end-to-end browser crawl, 2026-06-30)

What was verified

Not just install: a real crawl of a live page returning real markdown.

Commands run

# 1. package present
pip show crawl4ai            # Name: Crawl4AI, Version: 0.8.6

# 2. first crawl attempt FAILED with a real browser error:
#    BrowserType.launch: Executable doesn't exist at ...chromium-1194\chrome-win\chrome.exe
#    Root cause: Playwright browser version mismatch (Crawl4AI wanted chromium build 1194,
#    only 1223/1228 were installed). This is the #1 gotcha for this tool.

# 3. fix (the required post-install step):
crawl4ai-setup               # installs the matching chromium + patchright + inits the DB

# 4. real crawl, re-run:
python -c "
import asyncio, sys
if sys.platform.startswith('win'): asyncio.set_event_loop_policy(asyncio.WindowsProactorEventLoopPolicy())
from crawl4ai import AsyncWebCrawler
async def main():
    async with AsyncWebCrawler() as c:
        r = await c.arun('https://example.com')
        print('CRAWL_SUCCESS' if r.success else 'CRAWL_FAIL', 'status=%s markdown_chars=%s' % (r.status_code, len(r.markdown or '')))
asyncio.run(main())
"

Actual output

CRAWL_SUCCESS status=200 markdown_chars=166

Takeaway

pip install crawl4ai alone is not enough: a bare install cannot launch a browser. You must run crawl4ai-setup (or python -m playwright install chromium) after installing, or every crawl fails with a browser-not-found error. Once set up, Crawl4AI does a real render-and-extract to clean markdown, and is a free self-hosted alternative for the crawl-to-markdown job (pair it with your own proxies for volume).