cloakbrowser.dev website analysis
https://cloakbrowser.dev/
概要
- cloakbrowser
- ボリューム
- 16,240
- クリック単価
- $1.88
- 推定値
- $4,940
- cloak browser
- ボリューム
- 2,580
- クリック単価
- $1.97
- 推定値
- $1,130
- cloakbrowser浏览器
- ボリューム
- 410
- クリック単価
- --
- 推定値
- $390
- cloack browser
- ボリューム
- 350
- クリック単価
- --
- 推定値
- $250
- cloackbrowser
- ボリューム
- 300
- クリック単価
- --
- 推定値
- $220
| キーワード | ボリューム | CPC | 推定値 |
|---|---|---|---|
| cloakbrowser | 16,240 | $1.88 | $4,940 |
| cloak browser | 2,580 | $1.97 | $1,130 |
| cloakbrowser浏览器 | 410 | -- | $390 |
| cloack browser | 350 | -- | $250 |
| cloackbrowser | 300 | -- | $220 |
| チャンネル | 2026年6月 | 2026年7月 | 2026年8月 |
|---|---|---|---|
| perplexity.ai | 0.00% | 0.00% | 58.55% |
| chatgpt.com | 0.00% | 0.00% | 41.45% |
MRR
創業者またはチームが公開した MRR を公開情報から検索しています…
製品分析
製品のスナップショット
CloakBrowser
CloakBrowser is presented as a source-modified Chromium browser and wrapper for developers, automation teams, QA, browser agents, and profile operators who need Playwright/Puppeteer-style automation or persistent isolated browser profiles in environments where standard headless browsers trigger anti-bot checks. The site also presents CloakBrowser Manager as a self-hosted profile-management app powered by the same engine.
- カテゴリ
- Stealth browser automation / anti-detect browser / source-modified Chromium for Playwright, Puppeteer, Selenium, CDP, and profile management
- 位置決め
- Positioned as a real patched Chromium binary rather than a configuration tweak or JavaScript stealth injection, with source-level C++ fingerprint modifications, familiar automation APIs, self-hosted control, and optional Manager profiles for users who want anti-detect-style profile operations without cloud-hosted profiles.
- ビジネスモデル
- Freemium software subscription. The public site states that users can start free with one active browser session or one Manager profile at a time, then pay monthly plans based on concurrency. It also mentions managed CloakBrowser Cloud, usage-based browser-hour packages, Enterprise, and OEM/SaaS licensing as contact-based options.
Run browser automation through familiar Playwright, Puppeteer, Selenium, or CDP interfaces while reducing bot-detection signals associated with standard headless Chromium.
Collect or monitor public web pages using isolated, coherent browser sessions across sources, regions, or workers.
Test user journeys across browser environments, locales, timezones, viewports, and regional contexts without maintaining many physical devices.
Use a stronger underlying browser layer while keeping existing agent orchestration frameworks and browser-control APIs.
Maintain separated persistent browser profiles with independent cookies, fingerprints, proxies, locale settings, and session history.
解決された問題
- Standard Playwright, Puppeteer, or headless Chromium sessions can expose automation signals that trigger challenges, throttling, or blocks.
- JavaScript-injection or config-level stealth approaches are claimed to break or become inconsistent across Chrome updates.
- Running many workflows from one shared browser context can mix cookies, storage, hardware signals, locale, and browsing history.
- Teams may need persistent browser identities that can be reopened later without cross-profile leakage.
- Automation teams may want to keep existing Playwright/Puppeteer code instead of learning a new SDK.
中核的な問題点
- Anti-bot checks interrupt browser automation workflows.
- Headless browser fingerprints can be inconsistent with normal Chrome.
- Managing many isolated accounts or profiles can become costly or operationally fragile.
- Locale, timezone, proxy, and browser-history mismatches can distort research or testing results.
- Self-managed browser automation at scale requires binary management, profiles, Docker, and concurrency control.
価値提案
CloakBrowser claims to provide a drop-in Playwright/Puppeteer-compatible Chromium build with source-level fingerprint patches, giving teams more coherent browser behavior in anti-bot environments while keeping sessions, cookies, credentials, and infrastructure under the customer’s control.
コア機能
述べられた利点
- Claims source-level C++ browser modifications rather than JavaScript injection or launch-flag spoofing.
- Claims compatibility with standard Playwright and Puppeteer APIs, reducing migration effort.
- Self-hosted model is positioned as keeping sessions, page content, cookies, and credentials under user control.
- Manager pricing is positioned around concurrency with unlimited stored profiles, rather than per-profile limits.
- The same license key is stated to power both the developer library and Manager.
- The site claims the browser passes or improves outcomes on more than 30 detection checks, including examples such as Cloudflare Turnstile, reCAPTCHA v3, FingerprintJS, and BrowserScan.
- Manager app is stated to be open-source under MIT, though the binary itself remains licensed and restricted.
買収
Developer-led acquisition through the website, quickstart snippets, GitHub, PyPI, npm, Docker commands, blog/community content, Slack, and public community mentions.
変換
Conversion is driven by a free start option with one session/profile, no-card messaging, direct downloads for Manager, quickstart examples, published detection-test claims, and pricing CTAs for Solo, Team, Business, and Scale plans.
収益化
Monthly subscription plans are based on active or concurrent browser sessions; the site lists paid plans from Solo through Scale, plus contact-based Cloud, usage-based packages, Enterprise, and OEM/SaaS licensing.
トラストシグナル
- Public homepage includes a detection-results table claiming comparisons against stock Playwright across services such as reCAPTCHA v3, Cloudflare Turnstile, FingerprintJS, BrowserScan, and other detection checks.
- The site provides public install commands for pip, npm, Docker, and code examples for multiple languages.
- The Manager page states that the Manager app itself is open-source under MIT.
- The community page aggregates external posts, videos, press/tutorial references, Reddit comments, and social posts about CloakBrowser.
- The legal page provides Terms of Service, Privacy Policy, and Refund Policy content.
- The site publishes acceptable-use restrictions that prohibit abuse such as credential stuffing, automated account creation, unauthorized access, fraud, and unauthorized data collection.
リスクまたはギャップ
- The strongest detection-performance statements are website claims; the crawl does not include independent lab verification or reproducible test methodology.
- The legal terms explicitly state that CloakHQ does not guarantee avoidance of detection or successful operation on any particular website or anti-automation system.
- The product operates in a sensitive anti-bot/anti-detect category, creating compliance, platform-terms, and misuse risks for customers.
- CloakBrowser does not include CAPTCHA solving or built-in proxy rotation; users must bring their own proxies and handle workflow compliance.
- Self-hosted deployment means customers may need to manage infrastructure, concurrency, proxy routing, credentials, and operational security themselves.
- The public evidence does not show customer logos, audited security posture, uptime history, SOC 2/ISO certifications, or detailed enterprise support SLAs.
- The binary cannot be redistributed, resold, sublicensed, bundled, embedded, or offered as a browser-as-a-service without separate OEM/SaaS licensing.
製品の機会
- Expand buyer appeal by providing clearer independent benchmarks, reproducible detection-test methodology, and third-party validation.
- Develop enterprise trust materials such as security documentation, deployment guides, compliance guidance, and support SLAs.
- Deepen use-case packaging for QA, localization, brand verification, public-web intelligence, and AI browser-agent teams.
- Clarify managed CloakBrowser Cloud availability, onboarding, and pricing for teams that do not want to run infrastructure.
- Offer more operational templates for Docker, persistent CDP workers, proxy/locale alignment, and CI/CD browser automation.
- Strengthen responsible-use positioning with more explicit guidance for legitimate automation, platform terms compliance, and prohibited workflows.
分析されたページ
9
クロールの深さ
2
証拠メモ
クロールされたページ
分析の制限
- Analysis is based only on supplied public-page text from the crawl.
- Website claims about detection performance, source patches, community adoption, and product effectiveness were not independently verified.
- The crawl evidence does not include package registry metadata, GitHub repository details, source code review, customer contracts, customer logos, or usage analytics.
- The blog page evidence is truncated, so blog content could not materially inform the analysis.
- Pricing and product packaging are captured as presented in the supplied pages and may change outside this crawl.
ビルドのタイムライン
SEO
広告
アドセンス
テクノロジー
Security
Inspect this website’s public security configuration, certificates, DNS records, and browser protections.
HTTPS & redirects
SSL / TLS certificate
HTTP security headers
HSTS
Cookie security
Mixed content & script integrity
DNSSEC
CAA records
Email security
security.txt
CDN / WAF signals
These checks inspect public configuration at a point in time. They do not prove that a site is free of vulnerabilities. Only the initial HTML, response cookies, and exact hostname are inspected; JavaScript-created content, authenticated pages, active port scans, and external threat databases are outside this check. Missing or unverified evidence is shown separately from passed checks.

ソーシャル