Bright Data rents you exits and runs the unblocking for you. Wayfern gives you a long-lived anti-detect browser you drive over CDP. This is the honest split between them.
Most people reading an "alternatives" page do not actually need to move. Here is the split, stated as plainly as we can make it.
Stay with Bright Data if you are buying exits and someone else’s unblocking logic. Four networks — datacenter, ISP, residential and mobile — sit behind one account with one flag syntax, all reachable over HTTP, HTTPS and SOCKS5. Residential targeting goes finer than ours: country, US state, city, US ZIP and ASN, plus OS targeting on residential. Web Unlocker and the Browser API absorb IP rotation, fingerprint emulation, CAPTCHA solving and retries server-side, the SERP API is priced per 1,000 successful requests, and under default settings neither it nor Web Unlocker bills a failed request — a failed fetch is their problem, not your invoice. Wayfern has no equivalent to the datasets marketplace, the ready-made scrapers, or a mobile network, and if any of those is load-bearing for you, switching is a downgrade.
Move to Wayfern when the browser is the thing you need to control. Wayfern ships its own anti-detect browser and exposes it as a raw CDP endpoint: you connect Playwright or Puppeteer to a session, keep it open across as many domains as the run needs, and drive a real identity through the whole flow. Bright Data does not ship an anti-detect browser at all — its docs treat them as third-party products you plug proxies into — and its hosted Browser API sessions are capped at 60 minutes, time out after 5 idle minutes, and confine navigation to a single domain. If your work is a logged-in, multi-domain, hours-long flow, that cap is the reason to look elsewhere.
Bright Data’s unblocking products take a request and hand back a result. Wayfern hands you the browser and gets out of the way.
Web Unlocker is a request in, a result out: it manages IP rotation, browser fingerprints, sessions and CAPTCHA solving for you, and can force JavaScript rendering, return markdown, or return a PNG. That is genuinely less code to own — but the state between two calls is theirs, not yours, and you cannot reach into the page between them.
Wayfern inverts that. `wss://browser.wayfern.com/ws?token=…` is a Chrome DevTools Protocol endpoint; the session exists for exactly as long as your socket stays open. The gateway sets no fixed session ceiling and no per-domain restriction — a run ends when you disconnect, when your credits run out, or when the link’s concurrency cap refuses a new connection. Cookies, storage and the applied fingerprint stay put across every navigation in between.
The identity is not bolted on afterwards either. The browser exposes a `Wayfern.getFingerprint` / `refreshFingerprint` / `setFingerprint` CDP domain that the platform drives to generate and pin a per-target Windows, macOS or Linux identity, and the same anti-detect browsers back the Web Data endpoints. The trade is real: you own the automation logic that Bright Data would have run for you.
Bright Data cells come from their documentation and are listed in Sources at the bottom of this page. Where their docs do not say, this table says so instead of guessing.
| Capability | Wayfern | Them |
|---|---|---|
| First-party anti-detect browser | The browser is the product. A Wayfern CDP domain generates, applies and verifies a per-target Windows, macOS or Linux fingerprint. | None. The docs position anti-detect browsers as third-party tools you add Bright Data proxies to, and list 31 such products (GoLogin, Multilogin, AdsPower, Kameleo, Octo Browser…) as integration targets. |
| Hosted session limits | No fixed ceiling and no domain restriction: a session lives while the CDP socket is open and credits last. | Browser API: 60-minute maximum, automatic timeout after 5 idle minutes, and unlimited navigations only within the same domain. |
| Getting residential access | Self-serve. Add ?country=®ion=&city=&isp= to the /ws URL; codes are validated against the live catalog and a bad one closes the socket with 4007. | Registered companies only — personal-email accounts (Gmail, Outlook) are explicitly ineligible — with human-reviewed KYC returning within 48 hours, after which traffic outside the approved use case is refused with "Access denied". |
| Geo-targeting granularity | Country, region, city and ISP, plus a session id that holds one exit IP across a flow. | Country, US state, city, US ZIP and ASN on residential, plus Windows/MacOS/android OS targeting. Datacenter and ISP support country only. |
| Proxy networks on offer | One residential pool, attached to a browser session, plus a datacenter default exit whose egress is free. | Four: datacenter, ISP, residential and mobile, all supporting HTTP, HTTPS and SOCKS5. |
| Standalone proxy port | None. Exits attach to a browser session; there is no SOCKS5 endpoint to point other software at. | SOCKS5 on all four networks over port 22228, proxy-side DNS only; on residential and mobile it is restricted to HTTPS targets, a fixed allowed port list, and hostnames rather than raw IPs. |
| CAPTCHA solving | An explicit call you make: POST /api/captcha/solve, 1 credit debited on success. | Handled for you inside Web Unlocker and the Browser API, alongside rotation, fingerprint management and retries. |
| Ready-made data | None. Nine Web Data endpoints you call yourself: scrape, crawl, map, extract, screenshot, brand, styleguide, fonts, search. | A datasets marketplace advertising 250+ domains alongside 1000+ ready-to-use scrapers (vendor-stated catalog figures). |
| Search coverage | Google, Bing and DuckDuckGo, billed at 2 credits per provider page fetched. | SERP API lists seven engines — Google, Bing, DuckDuckGo, Yandex, Baidu, Yahoo, Naver — though dedicated query-parameter documentation exists only for Google and Bing. |
| Published per-GB proxy rate | Residential egress metered per byte at 320 credits/GB; datacenter egress is free. | No first-party per-GB or per-IP figure was readable while researching this page (not published) |
The shape of the bill matters more than the headline rate, and only one of these two rates is published.
Bright Data’s unblocking products bill for outcomes. The SERP API is priced per 1,000 successful requests with parsing, proxy management and unlocking included and no separate bandwidth fee, and neither it nor Web Unlocker bills failed requests under default settings. That is a genuinely good deal when unblocking is hard, and it is worth knowing the exception: switching on custom headers or cookies flips both products to billing 100% of requests, successful or not. Their docs also document a free tier — 5,000 credits a month, stated as a "~$7.50 value", covering Web Unlocker, SERP API, Web Scraper API and Scraper Studio, renewing on the 1st and not rolling over, plus a $2 one-time trial credit for the proxy and Browser API products.
We are not going to put a per-GB Bright Data number next to ours, because we could not read one from a first-party page while writing this. Third-party figures circulate and contradict each other. Check their pricing page directly and compare it against the rates below.
Wayfern meters one balance across everything. Browser sessions cost 8 credits per open-page hour, accrued per second with a 60-second minimum bill per page. Residential egress is pay-as-you-go at 320 credits/GB, metered per byte by a sidecar that samples every 5 seconds and terminates a session that can no longer pay; sessions without location parameters take the datacenter exit, whose egress costs nothing. The Web Data endpoints are per unit — 1 credit per scraped page, 1 credit per uncached source page, 5 credits per capture — and a captcha solve is 1 credit on success.
Plans are self-serve and the credits are the same currency for all of it: Developer at $19/mo for 1,520 credits and 20 concurrent sessions, Business at $99/mo for 9,000 credits and 100 concurrent sessions, Scale at $499/mo for 46,000 credits and 400 concurrent sessions. Credit packs roll over; plan grants do not.
Both are CDP endpoints, so the connection line is the only thing that really changes. What changes conceptually is where the exit is configured: Bright Data encodes targeting in the proxy username with flags like -country-xx, -city-xxxxx and -session-xxxxxx; Wayfern puts it in the query string of the connection itself, and the choice applies for the life of that session.
The other thing that changes is what you are allowed to do next. A Browser API session is one domain and at most 60 minutes; the second navigation below — to a different host, on the same session, with the same cookies and the same applied fingerprint — is the reason people move.
import { chromium } from 'playwright';
// Before — Bright Data Browser API: a hosted CDP endpoint, one domain per
// session, 60 minutes maximum, 5 idle minutes before it times out.
// const browser = await chromium.connectOverCDP(BRIGHTDATA_CDP_ENDPOINT);
// After — Wayfern: the same connectOverCDP call against a link token.
// Exit geo is connection state, not a username flag. session=<id> holds
// one residential IP for the whole run.
const browser = await chromium.connectOverCDP(
'wss://browser.wayfern.com/ws' +
'?token=<LINK_TOKEN>&country=US®ion=CA&city=<CITY>&session=run42'
);
const page = await browser.newPage();
await page.goto('https://example.com/login');
// ...log in, solve whatever needs solving...
// Same session, different host. Cookies, storage and the applied
// fingerprint all survive the hop.
await page.goto('https://partner.example.net/reports');
await browser.close(); // billing stops when the socket closesNo, and that is the main reason to read this page carefully. Wayfern sells browser sessions; residential egress attaches to a session through the /ws query string and is metered from the same credit balance, while the default datacenter exit is free. There is no standalone proxy port to point other software at, and no SOCKS5 endpoint. Bright Data offers four networks over HTTP, HTTPS and SOCKS5, so if you need an exit for something that is not a Wayfern browser, they are the ones who sell it.
No. Their documentation treats anti-detect browsers as third-party tools you add Bright Data proxies to, listing 31 of them — GoLogin, Multilogin, AdsPower, Kameleo, Octo Browser and others — as integration targets. What they do ship is the Browser API, a hosted cloud browser for Puppeteer, Playwright and Selenium with rotation, fingerprint management, CAPTCHA solving and retries handled server-side. Wayfern ships the anti-detect browser itself and hands you its CDP endpoint.
On Wayfern, yes: a session is bounded by your socket and your credit balance, not by a timer or a hostname, so a login on one domain and a fetch on another happen in the same browser with the same cookies and the same fingerprint. Bright Data’s Browser API documents the opposite constraint — unlimited navigations within one domain, a new session for a different domain, a 5-minute idle timeout and a 60-minute maximum.
Because we could not read Bright Data’s per-GB or per-IP proxy rate from a first-party page while researching this comparison, and the third-party numbers in circulation disagree with each other. Publishing one of them next to our own rate would be inventing a competitor’s price. Wayfern’s side is published and enforced by the code that bills you: 320 credits per GB of residential egress, free datacenter egress.
Concretely: the mobile and ISP networks, SOCKS5 access, ZIP-code, ASN and OS-level residential targeting, the datasets marketplace and the ready-made scrapers, and success-only billing on the unblocking APIs — under default settings, a failed Web Unlocker or SERP request costs you nothing there. You also take back ownership of the automation logic their managed products run for you. Those are real losses, and for plenty of workloads they outweigh anything on this page.
Browser sessions, the Web Data API, search and captcha solving all draw on one credit balance. Start on any of them.