[{"data":1,"prerenderedAt":117},["ShallowReactive",2],{"$fICfhrSh4nu0JmalwJXW1hXo88MQhLnnDEgc7evL_-a8":3},{"title":4,"date":5,"dateModified":6,"datePublished":7,"dateModifiedISO":7,"image":8,"content":9,"faq":10,"metaTitle":39,"metaDescription":40,"author":41,"authorBio":6,"authorLinkedin":6,"authorTitle":6,"authorPhoto":42,"lastReviewed":6,"noindex":43,"researchBasis":6,"category":44,"readingTime":45,"related":46,"prev":63,"next":66,"toc":69,"takeaways":111},"Camoufox Alternatives in 2026: 5 Options and the Licence Nobody Checks","20 Oct 2026",null,"2026-10-20","/img/news/camoufox-alternatives-2026.png","\u003Cp>\u003Cstrong>Hardly anyone leaves Camoufox because it fails to get in.\u003C/strong> They leave because of what it costs per page to get in — roughly 42 seconds per challenged page, 200MB+ per context, and a fleet that has to grow linearly with the catalogue.\u003C/p>\n\u003Cp>So an honest alternatives list is not a stealth leaderboard. It is a list sorted by the constraint that is actually binding on you, and for most teams that is throughput or licensing, not detection.\u003C/p>\n\u003Ch2 id=\"choose-before-you-read\">Choose Before You Read\u003C/h2>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Your reason for looking\u003C/th>\n\u003Cth>Go to\u003C/th>\n\u003C/tr>\n\u003C/thead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>Camoufox works but the crawl does not finish in time\u003C/td>\n\u003Ctd>\u003Ca href=\"#throughput-is-the-number-that-decides\">Throughput is the number\u003C/a>\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Legal asked what licence your scraper stack uses\u003C/td>\n\u003Ctd>\u003Ca href=\"#the-licence-axis-nobody-puts-in-these-lists\">The licence axis\u003C/a>\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>You are on Node.js and Camoufox is Python-first\u003C/td>\n\u003Ctd>\u003Ca href=\"#2-patchright-apache-20\">Patchright\u003C/a>\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>The target is Akamai, not Cloudflare\u003C/td>\n\u003Ctd>\u003Ca href=\"#4-curl_cffi-mit\">Drop the browser entirely\u003C/a>\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>You want the lowest memory per worker\u003C/td>\n\u003Ctd>\u003Ca href=\"#3-nodriver-agpl-30\">nodriver\u003C/a> — then read the licence section\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>You are not sure you should replace it at all\u003C/td>\n\u003Ctd>\u003Ca href=\"#camoufox-is-still-right-for\">Camoufox is still right for\u003C/a>\u003C/td>\n\u003C/tr>\n\u003C/tbody>\u003C/table>\n\u003Caside class=\"article__usecase-card\">\u003Cdiv class=\"article__usecase-label\">Related use case\u003C/div>\u003Ch3 class=\"article__usecase-title\">Compare ScrapeWise to your current tool\u003C/h3>\u003Cp class=\"article__usecase-blurb\">Side-by-side breakdowns against 20 scraping and price-monitoring platforms.\u003C/p>\u003Ca class=\"article__usecase-link\" href=\"/alternatives\">See how it works →\u003C/a>\u003C/aside>\u003Ch2 id=\"why-people-leave-camoufox-it-is-rarely-stealth\">Why People Leave Camoufox (It Is Rarely Stealth)\u003C/h2>\n\u003Cp>\u003Ca href=\"https://github.com/daijro/camoufox\">Camoufox\u003C/a> is a Firefox fork with anti-detection patches applied at the C++ engine level rather than injected as JavaScript after the page loads. That design is the reason it scores well — there is no patched property descriptor for a detection script to notice, because the behaviour was changed before the runtime existed.\u003C/p>\n\u003Cp>In our April 2026 benchmark it was the strongest of six approaches on Cloudflare: \u003Cstrong>91% on standard tiers and 78% on Enterprise\u003C/strong> with residential proxies. Nothing else in the table reached either figure.\u003C/p>\n\u003Cp>It was also the slowest and heaviest thing we ran. The three complaints that actually drive replacement:\u003C/p>\n\u003Cul>\n\u003Cli>\u003Cstrong>42.49 seconds\u003C/strong> average to clear a Cloudflare Turnstile challenge, which is about 85 challenged pages per hour per worker\u003C/li>\n\u003Cli>\u003Cstrong>200MB+ per browser context\u003C/strong>, so concurrency is bounded by RAM long before it is bounded by the target\u003C/li>\n\u003Cli>\u003Cstrong>A separate ~120MB binary\u003C/strong> to download and pin, which is a different operational story from \u003Ccode>pip install\u003C/code>\u003C/li>\n\u003C/ul>\n\u003Cp>None of those is a stealth failure. They are capacity facts, and they only matter at volume — which is exactly where price monitoring lives.\u003C/p>\n\u003Cdiv class=\"video-embed\">\u003Ciframe src=\"https://www.youtube-nocookie.com/embed/R_SlFcx5EuU\" title=\"How AI Agents Avoid Bot Detection with Camoufox\" loading=\"lazy\" allowfullscreen>\u003C/iframe>\u003C/div>\n\n\u003Ch2 id=\"the-five-alternatives-with-their-licences\">The Five Alternatives, With Their Licences\u003C/h2>\n\u003Cp>Checked directly against the GitHub API on \u003Cstrong>20 October 2026\u003C/strong>. Star counts and dates move; licences rarely do, and the licence is the column that turns out to matter.\u003C/p>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Tool\u003C/th>\n\u003Cth>Licence\u003C/th>\n\u003Cth>Stars\u003C/th>\n\u003Cth>Last push\u003C/th>\n\u003Cth>Replaces Camoufox when\u003C/th>\n\u003C/tr>\n\u003C/thead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>\u003Ca href=\"https://github.com/Kaliiiiiiiiii-Vinyzu/patchright\">Patchright\u003C/a>\u003C/td>\n\u003Ctd>Apache-2.0\u003C/td>\n\u003Ctd>4,920\u003C/td>\n\u003Ctd>15 Oct 2026\u003C/td>\n\u003Ctd>You need Chromium, Node.js, or speed\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https://github.com/seleniumbase/SeleniumBase\">SeleniumBase\u003C/a>\u003C/td>\n\u003Ctd>MIT\u003C/td>\n\u003Ctd>13,101\u003C/td>\n\u003Ctd>19 Oct 2026\u003C/td>\n\u003Ctd>You want UC mode plus a test framework\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https://github.com/ultrafunkamsterdam/nodriver\">nodriver\u003C/a>\u003C/td>\n\u003Ctd>\u003Cstrong>AGPL-3.0\u003C/strong>\u003C/td>\n\u003Ctd>4,869\u003C/td>\n\u003Ctd>8 Oct 2026\u003C/td>\n\u003Ctd>Memory per worker is the binding limit\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https://github.com/lexiforest/curl_cffi\">curl_cffi\u003C/a>\u003C/td>\n\u003Ctd>MIT\u003C/td>\n\u003Ctd>6,742\u003C/td>\n\u003Ctd>19 Oct 2026\u003C/td>\n\u003Ctd>The data is in the HTML — no browser needed\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>\u003Ca href=\"https://github.com/rebrowser/rebrowser-patches\">rebrowser-patches\u003C/a>\u003C/td>\n\u003Ctd>MIT\u003C/td>\n\u003Ctd>1,444\u003C/td>\n\u003Ctd>19 Oct 2026\u003C/td>\n\u003Ctd>You want to keep Playwright and patch its leaks\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Camoufox (the incumbent)\u003C/td>\n\u003Ctd>MPL-2.0\u003C/td>\n\u003Ctd>12,676\u003C/td>\n\u003Ctd>19 Oct 2026\u003C/td>\n\u003Ctd>—\u003C/td>\n\u003C/tr>\n\u003C/tbody>\u003C/table>\n\u003Ch3 id=\"1-seleniumbase-mit\">1. SeleniumBase (MIT)\u003C/h3>\n\u003Cp>The most under-rated option on the list, because people file it as a testing framework and stop reading. Its UC mode drives a real Chrome with the automation markers removed, and it clears interactive widgets that a vanilla driver cannot touch. If your team already writes Selenium, this is the shortest path from blocked to unblocked.\u003C/p>\n\u003Cp>Weakest on the hardest tier: Cloudflare Enterprise and DataDome will still beat it where Camoufox gets through.\u003C/p>\n\u003Ch3 id=\"2-patchright-apache-20\">2. Patchright (Apache-2.0)\u003C/h3>\n\u003Cp>A patched Chromium that is a drop-in for Playwright — in practice one import line. Measured \u003Cstrong>83% on Cloudflare standard and 69% on Enterprise\u003C/strong> with residential proxies, against Camoufox's 91/78. You give up eight to nine points and get back a much faster browser, Chromium instead of Firefox, and Node.js support.\u003C/p>\n\u003Cp>That trade is usually correct below Enterprise tier. Full detail in \u003Ca href=\"/blogs/camoufox-vs-patchright-2026\">Camoufox vs Patchright\u003C/a> and \u003Ca href=\"/blogs/what-is-patchright-guide-2026\">what Patchright actually patches\u003C/a>.\u003C/p>\n\u003Ch3 id=\"3-nodriver-agpl-30\">3. nodriver (AGPL-3.0)\u003C/h3>\n\u003Cp>The lightest option: \u003Cstrong>~80–120MB per instance\u003C/strong> against a browser context's 200MB+, because it drives unmodified Chrome over the DevTools Protocol with no framework in between and never fires the \u003Ccode>Runtime.enable\u003C/code> call that detection systems watch for.\u003C/p>\n\u003Cp>It is also the one that does nothing about behavioural scoring, so Cloudflare Enterprise, DataDome and Akamai still block it. See \u003Ca href=\"/blogs/nodriver-vs-playwright-2026\">nodriver vs Playwright\u003C/a> for where that ceiling sits.\u003C/p>\n\u003Cp>And it is AGPL-3.0. Read the next section before you adopt it.\u003C/p>\n\u003Ch3 id=\"4-curlcffi-mit\">4. curl_cffi (MIT)\u003C/h3>\n\u003Cp>Not a browser at all. It replays a real browser's TLS handshake at the socket level, which on some vendors is the only signal that matters. On Akamai Bot Manager it reached \u003Cstrong>87% — beating both stealth browsers\u003C/strong> — at roughly a quarter of the compute, because there is no Chromium process.\u003C/p>\n\u003Cp>On Cloudflare it is weaker than a stealth browser (79% standard, 52% Enterprise), so this is a per-vendor choice, not an upgrade. The mechanics are in \u003Ca href=\"/blogs/curl-cffi-tls-fingerprint-web-scraping-2026\">curl_cffi and TLS fingerprint impersonation\u003C/a> and the Akamai inversion is in \u003Ca href=\"/blogs/playwright-akamai-bot-manager-2026\">Playwright vs Akamai Bot Manager\u003C/a>.\u003C/p>\n\u003Ch3 id=\"5-rebrowser-patches-mit\">5. rebrowser-patches (MIT)\u003C/h3>\n\u003Cp>The conservative option: keep Playwright or Puppeteer and patch the known CDP leaks rather than adopt a new runtime. Smallest migration cost on this list and the smallest gain. Sensible when you have a large existing Playwright codebase and a detection problem that is narrower than it feels.\u003C/p>\n\u003Caside class=\"article__inline-cta\">\u003Cp class=\"article__inline-cta-text\">Try ScrapeWise on your own URL. \u003Cstrong>Your first 5 requests are free.\u003C/strong>\u003C/p>\u003Ca class=\"article__inline-cta-btn\" href=\"https://portal.scrapewise.ai/register\" target=\"_blank\" rel=\"noopener\">Start Free →\u003C/a>\u003C/aside>\u003Ch2 id=\"the-licence-axis-nobody-puts-in-these-lists\">The Licence Axis Nobody Puts in These Lists\u003C/h2>\n\u003Cp>Every roundup of Camoufox alternatives ranks on stealth and speed. We have not found one that prints the licence column, and it is the column most likely to get a decision reversed after the fact.\u003C/p>\n\u003Cp>\u003Cstrong>\u003Ca href=\"https://www.gnu.org/licenses/agpl-3.0.en.html\">AGPL-3.0\u003C/a>\u003C/strong> — nodriver's licence — extends copyleft to network use. Where a normal GPL obligation is triggered by distributing a binary, the AGPL's section 13 is triggered by letting users interact with the software \u003Cstrong>over a network\u003C/strong>. If scraped data reaches your customers through a product that embeds it, whether that creates an obligation is a real question with a real answer, and the answer is your counsel's to give, not a blog's.\u003C/p>\n\u003Cp>\u003Cstrong>\u003Ca href=\"https://www.mozilla.org/en-US/MPL/2.0/\">MPL-2.0\u003C/a>\u003C/strong> — Camoufox's licence — is file-level copyleft. Modify Camoufox's own files and those files stay open; your surrounding code does not. This is why Camoufox is generally unproblematic inside commercial products.\u003C/p>\n\u003Cp>\u003Cstrong>Apache-2.0 and MIT\u003C/strong> — Patchright, SeleniumBase, curl_cffi, rebrowser-patches — are permissive. Apache-2.0 additionally grants a patent licence, which some legal teams specifically ask for.\u003C/p>\n\u003Cp>None of this is legal advice and we are not qualified to give it. The point is narrower and worth saying plainly: \u003Cstrong>of the five alternatives, exactly one has a licence that can turn an infrastructure choice into a product decision, and it is routinely recommended without that being mentioned.\u003C/strong> Check it before you build on it, not after.\u003C/p>\n\u003Ch2 id=\"throughput-is-the-number-that-decides\">Throughput Is the Number That Decides\u003C/h2>\n\u003Cp>If you are replacing Camoufox for capacity rather than access, this is the arithmetic to run.\u003C/p>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>\u003C/th>\n\u003Cth>Camoufox\u003C/th>\n\u003Cth>Patchright\u003C/th>\n\u003Cth>nodriver\u003C/th>\n\u003Cth>curl_cffi\u003C/th>\n\u003C/tr>\n\u003C/thead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>Memory per worker\u003C/td>\n\u003Ctd>200MB+\u003C/td>\n\u003Ctd>~200MB\u003C/td>\n\u003Ctd>\u003Cstrong>~80–120MB\u003C/strong>\u003C/td>\n\u003Ctd>negligible\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Cloudflare Std / Enterprise\u003C/td>\n\u003Ctd>\u003Cstrong>91% / 78%\u003C/strong>\u003C/td>\n\u003Ctd>83% / 69%\u003C/td>\n\u003Ctd>blocked\u003C/td>\n\u003Ctd>79% / 52%\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Akamai v4\u003C/td>\n\u003Ctd>83%\u003C/td>\n\u003Ctd>74%\u003C/td>\n\u003Ctd>blocked\u003C/td>\n\u003Ctd>\u003Cstrong>87%\u003C/strong>\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Challenged pages per hour per worker\u003C/td>\n\u003Ctd>~85\u003C/td>\n\u003Ctd>faster, not measured per challenge\u003C/td>\n\u003Ctd>n/a\u003C/td>\n\u003Ctd>no challenge step\u003C/td>\n\u003C/tr>\n\u003C/tbody>\u003C/table>\n\u003Cp>Work it from the brief, not the tool. Four competitors, 5,000 SKUs, daily, is 20,000 pages. At a 20% challenge rate that is 4,000 challenges, and at 85 per hour per worker it is 47 worker-hours a day. On a cheaper tool that passes less often you will run more retries; on a heavier one you will run more workers. Both end up as a line on an invoice.\u003C/p>\n\u003Cp>The configuration most teams land on is not a single tool at all: \u003Cstrong>route per domain.\u003C/strong> Cheapest approach by default, measure block rate per target rather than site-wide, promote only the failing handful to the heavy browser with its own concurrency cap. A site-wide 8% failure figure usually hides one domain failing 90% of the time while thirty sit at zero.\u003C/p>\n\u003Ch2 id=\"camoufox-is-still-right-for\">Camoufox Is Still Right For\u003C/h2>\n\u003Cp>Worth stating, because "alternatives" articles tend to conclude that you should switch:\u003C/p>\n\u003Cul>\n\u003Cli>\u003Cstrong>Cloudflare Enterprise targets.\u003C/strong> 78% is the highest number we measured. Nine points over Patchright is a lot of retries.\u003C/li>\n\u003Cli>\u003Cstrong>High-value, low-frequency pages.\u003C/strong> If you need 500 pages a day from a hard target, 42 seconds each is 6 hours and nobody notices.\u003C/li>\n\u003Cli>\u003Cstrong>Firefox-shaped fingerprints specifically.\u003C/strong> The whole reason it wins on Cloudflare is that it is not in the Chromium family.\u003C/li>\n\u003Cli>\u003Cstrong>Commercial products with a careful legal review.\u003C/strong> MPL-2.0 is the easiest licence on this list to live with after AGPL is ruled out.\u003C/li>\n\u003C/ul>\n\u003Cp>Switching away from the tool with the highest pass rate because it is slow, then discovering the replacement needs three retries per page, is a round trip a lot of teams have made.\u003C/p>\n\u003Ch2 id=\"what-none-of-them-replace\">What None of Them Replace\u003C/h2>\n\u003Cp>All six tools here hand you a way to fetch a page. None hands you data.\u003C/p>\n\u003Cp>What stays yours regardless of the choice: writing selectors and rewriting them after every redesign, buying and rotating proxies, session and cookie management, retry logic, storage, change detection, and the matching layer that turns rows from six retailers into one comparable table. The fetching tool is the part of this problem that attracts the most attention and consumes the least ongoing time.\u003C/p>\n\u003Cp>If what you want is \u003Cstrong>competitor prices and stock levels arriving in your systems\u003C/strong> rather than a browser fleet to keep alive, the Camoufox-versus-everything question is the wrong one to be answering. \u003Ca href=\"/\">ScrapeWise\u003C/a> runs the browsers, the proxies and the anti-bot escalation and delivers structured rows. No plan and no compute units: a balance you top up, and a published charge per delivered page — EUR 0.15 per 1,000 for plain requests, EUR 0.75 with rendering, EUR 1.50 through the residential tier and EUR 3.75 for render plus residential. New accounts get 5 free requests and the balance never expires.\u003C/p>\n\u003Cp>The honest boundary: we are a data feed, not an automation framework. If you are building something interactive, keep the browser and keep the library.\u003C/p>\n\u003Cp>You can also call \u003Ca href=\"/scrapers\">ready-made scraping API endpoints\u003C/a> directly, or see how this feeds \u003Ca href=\"/use-cases/competitor-price-tracking\">competitor price tracking\u003C/a>.\u003C/p>\n\u003Ch2 id=\"so-which-should-you-pick\">So Which Should You Pick?\u003C/h2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Stay on Camoufox\u003C/strong> — Cloudflare Enterprise, high-value pages, volume you can bound. It is still the best pass rate we measured.\u003C/li>\n\u003Cli>\u003Cstrong>Patchright\u003C/strong> — you want most of the access for materially more speed, on Chromium, with Node.js available.\u003C/li>\n\u003Cli>\u003Cstrong>SeleniumBase\u003C/strong> — your team already writes Selenium and the targets are mid-tier.\u003C/li>\n\u003Cli>\u003Cstrong>curl_cffi\u003C/strong> — the target is Akamai, or the data is already in the HTML. The biggest cost saving on the list.\u003C/li>\n\u003Cli>\u003Cstrong>nodriver\u003C/strong> — memory is the binding constraint and the targets are light. Settle the AGPL question first.\u003C/li>\n\u003Cli>\u003Cstrong>rebrowser-patches\u003C/strong> — a large existing Playwright codebase you do not want to migrate.\u003C/li>\n\u003Cli>\u003Cstrong>None of them\u003C/strong> — you want rows on a schedule and would rather not run a browser fleet to get them.\u003C/li>\n\u003C/ul>\n\u003Cp>\u003Ca href=\"/pricing\">See pricing\u003C/a>, or compare the two incumbents head to head in \u003Ca href=\"/blogs/camoufox-vs-playwright-2026\">Camoufox vs Playwright\u003C/a>.\u003C/p>\n",{"title":11,"description":12,"badge":13,"benefits":14},"Frequently asked questions","The five realistic Camoufox alternatives, their software licences, and the throughput arithmetic that decides between them","FAQ",[15,18,21,24,27,30,33,36],{"title":16,"description":17},"What are the best alternatives to Camoufox in 2026?","Five realistic options, each winning on a different axis. Patchright, a patched Chromium that is close to a drop-in for Playwright, when you need speed, Chromium or Node.js. SeleniumBase, when your team already writes Selenium and wants UC mode plus a test framework. nodriver, when memory per worker is the binding constraint. curl_cffi, when the data is already in the HTML and no browser is needed at all. rebrowser-patches, when you have a large existing Playwright codebase and want to patch its leaks rather than migrate. Which one is right depends on whether your constraint is throughput, licensing, language or the specific anti-bot vendor in front of the target.",{"title":19,"description":20},"Why do teams replace Camoufox if it has the best pass rate?","Almost never because of stealth. Camoufox was the strongest of six approaches in our April 2026 benchmark, at 91% on Cloudflare standard tiers and 78% on Enterprise with residential proxies. It is also the slowest and heaviest thing we ran: roughly 42.49 seconds to clear a Turnstile challenge, which is about 85 challenged pages per hour per worker, over 200MB per browser context, and a separate binary of around 120MB to download and pin. Those are capacity facts rather than detection failures, and they only start to hurt at volume, which is exactly where price monitoring lives.",{"title":22,"description":23},"What licence is nodriver, and does it matter for commercial use?","nodriver is licensed AGPL-3.0, which we verified directly against the GitHub API on 20 October 2026. The AGPL extends copyleft to network use: where an ordinary GPL obligation is triggered by distributing a binary, the AGPL's section 13 is triggered by letting users interact with the software over a network. If scraped data reaches your customers through a product that embeds that code, whether an obligation arises is a genuine question and the answer belongs to your legal counsel, not to a blog post. The narrow point worth making is that of the five common alternatives exactly one carries this licence, and most roundups recommend it without mentioning the licence at all.",{"title":25,"description":26},"What licences do the other Camoufox alternatives use?","Checked against the GitHub API on 20 October 2026: Camoufox itself is MPL-2.0, Patchright is Apache-2.0, SeleniumBase is MIT, curl_cffi is MIT, and rebrowser-patches is MIT. MPL-2.0 is file-level copyleft, meaning modifications to Camoufox's own files stay open while your surrounding code does not, which is why Camoufox is generally unproblematic inside commercial products. Apache-2.0 and MIT are permissive, and Apache-2.0 additionally grants a patent licence that some legal teams specifically ask for. None of this is legal advice; it is the column that every alternatives list leaves out.",{"title":28,"description":29},"Is Patchright a good replacement for Camoufox?","Usually yes below Cloudflare Enterprise tier. Patchright measured 83% on Cloudflare standard and 69% on Enterprise with residential proxies, against Camoufox at 91% and 78%. You give up eight to nine points and get back a materially faster browser, Chromium instead of Firefox, Node.js support, and a migration that is in practice one import line. Where that trade stops working is Enterprise and the hardest DataDome deployments, where nine points of pass rate turns into a large number of retries and the cheaper tool ends up costing more per delivered page than the expensive one.",{"title":31,"description":32},"Is there a faster alternative to Camoufox than another browser?","Yes, and on some vendors it is also more effective. curl_cffi is not a browser: it replays a real browser's TLS handshake at the socket level. Against Akamai Bot Manager v4 it reached 87%, beating both stealth browsers we tested, at roughly a quarter of the compute because there is no Chromium process at all. Against Cloudflare it is weaker than a stealth browser at 79% on standard and 52% on Enterprise. So it is a per-vendor choice rather than a straight upgrade, and the correct configuration for most teams is to route per domain rather than standardise on one tool.",{"title":34,"description":35},"When should you stay on Camoufox?","When the target is Cloudflare Enterprise, where 78% is the highest figure we measured and nine points over Patchright is a lot of avoided retries. When the pages are high-value and low-frequency, because 500 pages a day at 42 seconds each is six hours and nobody notices. When the Firefox-shaped fingerprint is specifically the reason you are getting through, which is the whole mechanism behind its Cloudflare advantage. Switching away from the highest pass rate because it is slow, then discovering the replacement needs three retries per page, is a round trip a lot of teams have made.",{"title":37,"description":38},"How do you choose between these tools at scale?","Work from the brief rather than the tool. Four competitors at 5,000 SKUs refreshed daily is 20,000 pages; at a 20% challenge rate that is 4,000 challenges, and at 85 per hour per worker that is 47 worker-hours a day. A cheaper tool that passes less often spends the difference on retries and a heavier one spends it on workers, so both land on the invoice. The configuration most teams settle on is not a single tool: route per domain, measure block rate per target rather than site-wide, and promote only the failing handful to the heavy browser. A site-wide 8% failure figure usually hides one domain failing 90% of the time while thirty sit at zero.","Camoufox Alternatives 2026: 5 Options Compared","Five Camoufox alternatives compared on throughput per worker and software licence. One of the five is AGPL-3.0, and most roundups never mention it.","Siim Brazier","/img/team/siim.jpg",false,"Alternatives",9,[47,53,58],{"slug":48,"title":49,"image":50,"date":51,"category":44,"excerpt":52},"apify-vs-bright-data-web-scraping-2026","Apify vs Bright Data: Which Should You Choose in 2026?","/img/news/apify-vs-bright-data-web-scraping-2026.png","24 Sep 2026","Apify vs Bright Data in 2026: Apify meters compute units, Bright Data meters proxy GB, so the same job costs opposite amounts. Which one your own pages suit.",{"slug":54,"title":55,"image":56,"date":51,"category":44,"excerpt":57},"apify-vs-zyte-web-scraping-2026","Apify vs Zyte: Which Web Scraping Platform Should You Choose in 2026?","/img/news/apify-vs-zyte-web-scraping-2026.png","Apify vs Zyte in 2026: consumption-based compute units against $0.13–$1.27 per 1,000 successful responses, and who fixes the parser when a retailer redesigns.",{"slug":59,"title":60,"image":61,"date":51,"category":44,"excerpt":62},"bright-data-vs-oxylabs-web-scraping-2026","Bright Data vs Oxylabs: Which Should You Choose in 2026?","/img/news/bright-data-vs-oxylabs-web-scraping-2026.png","Bright Data vs Oxylabs in 2026: why IP-pool counts don't decide it, the cost-per-successful-page test that does, and where each vendor is genuinely ahead.",{"slug":64,"title":65},"curl-cffi-tls-fingerprint-web-scraping-2026","curl_cffi and TLS Fingerprint Impersonation for Web Scraping (2026)",{"slug":67,"title":68},"bypass-kasada-bot-protection-2026","Bypassing Kasada in 2026: The Token Economics Nobody Prices In",[70,74,77,80,84,87,90,93,96,99,102,105,108],{"level":71,"text":72,"id":73},2,"Choose Before You Read","choose-before-you-read",{"level":71,"text":75,"id":76},"Why People Leave Camoufox (It Is Rarely Stealth)","why-people-leave-camoufox-it-is-rarely-stealth",{"level":71,"text":78,"id":79},"The Five Alternatives, With Their Licences","the-five-alternatives-with-their-licences",{"level":81,"text":82,"id":83},3,"1. SeleniumBase (MIT)","1-seleniumbase-mit",{"level":81,"text":85,"id":86},"2. Patchright (Apache-2.0)","2-patchright-apache-20",{"level":81,"text":88,"id":89},"3. nodriver (AGPL-3.0)","3-nodriver-agpl-30",{"level":81,"text":91,"id":92},"4. curl_cffi (MIT)","4-curlcffi-mit",{"level":81,"text":94,"id":95},"5. rebrowser-patches (MIT)","5-rebrowser-patches-mit",{"level":71,"text":97,"id":98},"The Licence Axis Nobody Puts in These Lists","the-licence-axis-nobody-puts-in-these-lists",{"level":71,"text":100,"id":101},"Throughput Is the Number That Decides","throughput-is-the-number-that-decides",{"level":71,"text":103,"id":104},"Camoufox Is Still Right For","camoufox-is-still-right-for",{"level":71,"text":106,"id":107},"What None of Them Replace","what-none-of-them-replace",{"level":71,"text":109,"id":110},"So Which Should You Pick?","so-which-should-you-pick",[112,113,114,115,116],"Almost nobody leaves Camoufox because of stealth. They leave because of throughput: roughly 85 challenged pages per hour per worker, at 200MB+ per context.","nodriver is licensed AGPL-3.0. If you serve scraped data over a network from a product, that is a legal question for your counsel, not a footnote — and no alternatives roundup we found mentions it.","Camoufox is MPL-2.0, Patchright Apache-2.0, SeleniumBase MIT, curl_cffi MIT, rebrowser-patches MIT. Checked against the GitHub API on 20 October 2026.","The fastest alternative is not a browser. On Akamai a TLS-impersonating HTTP client beat every stealth browser we tested, at roughly a quarter of the compute.","Camoufox is still the right answer on Cloudflare Enterprise, where it passed 78% against Patchright's 69% in our April 2026 run. It wins where it wins; it is not a default.",1792524068781]