Updated November 2026
If you scrape the web, Cloudflare will block you. That's a fact in 2026. Every serious scraper eventually needs a bypass API. But which one? This guide compares the three most common choices with real numbers — not marketing copy.
Cloudflare now serves over 20% of all websites and protects most of the top 10,000. Its Turnstile challenge, TLS fingerprinting and JavaScript challenges can't be solved with plain requests. You either need a headless browser or an API that handles it for you.
Running your own headless browsers is expensive: 2 GB of RAM per Chrome instance, proxies, retries, captcha solvers. For most teams, a bypass API is cheaper than DIY.
| Feature | CF Bypass | ZenRows | ScraperAPI |
|---|---|---|---|
| Price per success | $0.03 | ~$0.05–0.10 | ~$0.10–0.15 |
| Price per 404 | $0.01 | Full price | Full price |
| Refund on failure | Yes | No | No |
| Response format | Raw HTML / binary (transparent proxy) | Raw HTML or JSON | Raw HTML or JSON |
| Original Content-Type preserved | Yes | Usually | Usually |
| Pay-as-you-go crypto | USDT/USDC (TRC-20, BEP-20) | Card only | Card only |
| Free trial | Yes, via small deposit | 1,000 requests | 5,000 requests |
| Minimum monthly commitment | None | $49/mo | $49/mo |
Best for indie developers, small scraping teams and anyone who wants to pay only for results. If your request fails, you don't pay. If the target returns 404, you pay $0.01 instead of full price. Crypto payment means no card required.
It returns raw bytes — exactly what the origin server sent. Your existing code that expects HTML works without changes:
import requests
resp = requests.post(
"https://cloudflare-bypass.shop/v1/solve",
headers={"X-API-Key": "YOUR_KEY"},
json={"url": "https://example.com"},
timeout=200
)
print(resp.status_code, len(resp.text))
ZenRows shines when your target is a heavy SPA and you need rendered DOM, screenshots, or specific JS instructions. Their pricing assumes you'll scrape a lot per month, so the subscription makes sense at scale. But per-success cost is higher and failures still bill you.
ScraperAPI has the widest feature set (structured data endpoints for Amazon, Google, etc.) and the best enterprise support. If you need a single vendor for everything and don't care about cost, it's a solid choice. For indie work, the $49/mo minimum usually kills it.
Most bypass APIs bill you for attempts, not successes. On a site that returns 404 for 30% of URLs, that triples your effective price. CF Bypass's model — refund on failure, $0.01 on 404 — was designed to fix this. Over a month, this often beats subscription APIs by 60–70%.
This trips up a lot of people. A bypass API that returns a "success" page may just be a Cloudflare interstitial. Check these:
If all four are true, you got the real page. Don't second-guess it.
Can I use a bypass API for commercial scraping?
Usually yes, but you're responsible for the target site's terms of service and local law. Bypass APIs just fetch pages; they don't grant legal rights.
Will Cloudflare block the bypass API itself?
The good ones rotate residential proxies and browsers on every request, so they stay ahead. CF Bypass uses a cascading approach: if one method fails, it retries with another before charging you.
Do I need to manage proxies?
No. The API handles proxies, browsers, retries and captchas internally.
Related: Python scraping guide · ZenRows alternatives · Scrapy with Cloudflare