Compare / other services
cdpfleet vs ScrapingBee
ScrapingBee's web scraping API (one HTTPS request per page, rendered in a headless browser, 1–75 credits by proxy tier, no browser to connect to) and its search and site APIs, against twelve real browsers you drive on flat threads and cdpfleet Functions. Prices, limits and features side by side.
ScrapingBee sells a web scraping API paid in credits from a monthly balance: you send one HTTPS request per URL and get the page back — rendered in a headless browser by default, through classic, premium or stealth proxies, after optional browser steps (js_scenario), as HTML, Markdown, text, JSON, extracted data or a screenshot — plus search and site APIs (Google, Fast Search, Amazon, Walmart, YouTube, ChatGPT, Gemini). There is no browser you connect to: no CDP or Playwright endpoint is documented, and the browser lives only for the length of one request. So this page compares ScrapingBee's API with cdpfleet Functions (one HTTP call per task, plus jobs and monitors), and notes where you'd need a session instead. cdpfleet is a fleet of twelve real desktop browsers — Google Chrome, Microsoft Edge, Brave, Opera, Naver Whale, Yandex Browser, Chromium, Patchright, Rebrowser, Camoufox, Firefox and WebKit — launched with one HTTPS call and driven over the Playwright protocol or CDP from five languages, through any proxy you choose, on threads you buy.
| cdpfleet | ScrapingBee | |
|---|---|---|
| Browsers | 12 — Google Chrome, Microsoft Edge, Brave, Opera, Naver Whale, Yandex Browser, Chromium, Patchright, Rebrowser, Camoufox, Firefox, WebKit. Chrome with stable, beta and dev channels and pinned previous majors; headful by default. | A headless browser behind the API — "Chrome rendering" on the Screenshot page; the version isn't published and you can't choose another. device=desktop or mobile, window_width/window_height (default 1920×1080). |
| How you connect | Playwright protocol (connect(wsUrl), client 1.60) from Node.js, Python, Java, C# and Go — or "cdp": true for a cdpUrl (Puppeteer, connectOverCDP). SDKs for Node.js, Python and Go (C# and Java build from source). | Not a browser you connect to. One request per page — GET https://app.scrapingbee.com/api/v1?url=… with options as query parameters — from any HTTP client, their SDKs or a CLI. A proxy mode (proxy.scrapingbee.com:8886/8887, SOCKS5 on 8888) puts the same API behind a proxy address for your own tools; each request through it is billed like an API call, and the docs recommend turning JavaScript rendering off there. |
| Session length | Up to 24 h (shared threads: also capped by your remaining daily minutes). Idle timeout 120 s by default, up to 10 min on shared threads and 24 h on dedicated. | One request: timeout 1–140 s (default 140 s), and the whole js_scenario must finish within 40 s. The docs' introduction also warns of "the maximum 30-second timeout", which doesn't match the 140 s parameter. session_id routes requests through the same IP for 5 minutes; keeping cookies or storage between requests isn't documented — you pass cookies yourself. |
| Browser time | Free Community plan: 2 threads, 60 browser-minutes per thread a day, 15-minute sessions, proxies included. Per second, as thread-seconds. Shared: $0.02–$0.05 per browser-hour at full use; dedicated $0.01/hour ($8.50/month for a browser around the clock). A headful session takes 2 threads. | Not sold — you pay per successful request, whatever it takes: classic proxy 1 credit without JavaScript, 5 with (the default); premium proxy 10 / 25; stealth proxy 75 (JavaScript only); AI extraction +5. Billed: 200, 404, 410 (and 413 in the status table); not billed: 400, 401, 403, 429, 500. mode=auto tries 1 → 5 → 10 → 25 → 75 credits until one works and charges 0 if all fail. |
| Traffic | Free through your own proxies — HTTP, HTTPS or SOCKS5, single, pools or per-host rules, with live swap. | Not metered: a request costs the same whatever the page weighs (images and CSS are blocked by default, block_resources). Your own proxy works with own_proxy; the docs don't say whether it changes the credit cost. |
| Proxies | Bring your own HTTP, HTTPS or SOCKS5 proxies — single, pool, or per-host rules (proxy_rules) — swap a running session's proxy live, see traffic per destination host and per rule; traffic through them is free. Or the cdpfleet residential proxy: 4 tiers from $0.50 to $5/GB from one prepaid balance ($5 free credit), mixable per host — e.g. Tier 4 for CDN assets, Tier 1 for pages. A launch without a proxy uses the free Community proxy — best effort, no guarantees. | Rotating classic and premium (residential) pools and a stealth pool for "the hardest" sites; country_code targeting — 200+ countries, 40+ on classic proxies (pricing page); own_proxy for your own. Stealth doesn't support custom headers, cookies, timeout or infinite_scroll. |
| Plans | Free Community plan: 2 threads, 60 browser-minutes per thread a day, 15-minute sessions, proxies included. Paid from $7.50/month (5 shared threads, 60–480 browser-minutes a day per thread); dedicated threads for 24/7 use at $8.50/thread/month. Pricing. | Free: 1,000 credits, no card. Hobby $19/month (75,000 credits, 25 concurrent), Freelance $49 (250,000, 50), Startup $99 (1 million, 100), Business $249 (3 million, 200), Business + $599 (8 million, 400); Enterprise $999 (14 million, 500) to $5,799 (120 million, 900). Credits expire monthly, no rollover; prices exclude VAT. Priority support, account manager and team management from Business. |
| Concurrency | The threads you buy: 5 to 10,000 (more on request), 2 on the free Community plan; each session takes 1–4 threads by type. | 25 to 900 concurrent requests by plan; over the limit you get a 429 (not billed). |
| Stealth | Real browsers by default (genuine Chrome, Edge, Brave…), headful by default, and three stealth builds — Patchright, Rebrowser, Camoufox (compare them). | Proxy tiers rather than browser choice: premium and stealth pools, mode=auto to escalate. The browser's fingerprint isn't documented, and we haven't measured how its rendered requests look to a fingerprinting page. |
| Captcha solving | Not built in. Hand over to a person with live view, or use your own solver through Playwright. Warmup never solves a CAPTCHA — it leaves and comes back later. | No CAPTCHA parameter in the API docs; the MCP server page says the API "handles proxies, CAPTCHAs, and JavaScript rendering". How often a given site still blocks would need a measurement. |
| Watching and keeping sessions | Live view with human takeover — Watch / Take control in the dashboard, or a link from the API — and Launch interactive: open any browser on a start page straight into live view from the dashboard's code generator (start_url over the API). Session recordings as MP4 (480p free, 720p or 1080p on a plan), and JPEG frames to your code with page.screencast. Persistent profiles keep cookies, local storage and IndexedDB between sessions (paid plans); Playwright storageState; keep_alive to reconnect to the same browser on dedicated threads. | Nothing to watch or keep: no live view, recordings or profiles are documented — each response is the result. session_id keeps one IP for 5 minutes. |
| Scraping in one call | Functions: screenshot, PDF, Markdown, scrape (HTML, Markdown, text, links, screenshot, XHR calls, cookies; wait_for, actions, extract by CSS selector), map, geo compare and a live tab — in a pooled browser, or in a browser and device of your choice on paid plans. Billed like a session (one thread while the call runs, one per country for geo compare); the Community proxy and your own proxies cost nothing extra. Free tools at /functions. | The HTML API: HTML (rendered or original source), Markdown, text, or JSON with the XHR calls, cookies, iframes and evaluate results; js_scenario (click, fill, wait, wait_for, scroll, infinite_scroll, evaluate; strict mode), wait, wait_for, wait_browser; extract_rules (CSS/XPath) or ai_query/ai_extract_rules (+5 credits). 1,000 rendered pages cost about $1.27 (Hobby) to $0.37 (Business +), $6.33–$1.87 with premium proxies and $19.00–$5.62 with stealth proxies — our arithmetic from the list prices. A per-page cost for a Function depends on how long the page takes, which we haven't measured against these. |
| Crawling and bulk | Jobs on paid plans: batch (any number of URLs), crawl and map, with no size or time limit — runs continue past the 24-hour session limit by themselves; results as JSON lines or a zip archive, kept 14 days; stop and resume; signed webhooks. Plus watch (live XHR/WebSocket/DOM events), warmup and monitors (scheduled change detection with diffs). | No crawl or batch endpoint in the API docs. The scrapingbee-cli (pip) has batch and crawl commands — same-domain links, a sitemap or your own Scrapy spider, with Scrapy under the hood — that call the API once per page, up to your plan's concurrency. |
| Search and site APIs | None — no search-engine or per-site APIs; a scrape or a session reads whatever page you point it at. | Google Search API (10 credits a search with "light request", 15 otherwise), Fast Search API (10 credits), and APIs for Amazon, Walmart, YouTube, ChatGPT, Gemini and employee search. |
| Screenshots and PDFs | Screenshot (PNG/JPEG, viewport or full page) and PDF Functions — a text PDF from a Chromium-based browser. | screenshot=true on the HTML API: viewport, full page or a CSS selector, optionally with the HTML in the same call; billed like any request. HTML-to-PDF is listed as "coming soon". |
| Agent tooling | A hosted MCP server at cdpfleet.com/mcp — add it in Claude, ChatGPT or Gemini and sign in with OAuth — or npx cdpfleet-mcp; SDKs for Node.js, Python and Go. | A remote MCP server (mcp.scrapingbee.com/mcp) authenticated with an API-key header — for Cursor, Claude Code or Claude Desktop (header auth there in beta); ChatGPT's in-app connectors aren't supported, only the OpenAI API. Integrations for n8n, Make, Zapier and LangChain; the CLI. |
| Examples | 89 cases run on production in five languages, all on GitHub with their output. | Code for every parameter in cURL, Python, Node.js, Java, Ruby, PHP and Go in the docs, and many blog tutorials. |
ScrapingBee facts from its public pricing page and documentation, checked on 4 November 2026; list prices in USD. Tell us if something is out of date.
Which one, when
Pick ScrapingBee if…
- You fetch pages rather than drive a browser: one GET per URL returns rendered HTML, Markdown, JSON or a screenshot, with proxies, retries and rendering handled.
- You want Google, Amazon, Walmart, YouTube or ChatGPT data, AI extraction and pages from one credit balance.
- Your pages are heavy but each is one short visit: a request costs the same whatever it weighs, failed requests aren't billed, and
mode=autofinds the cheapest proxy tier that works.
Pick cdpfleet if…
- You need a browser you drive: logins, multi-step flows, downloads and uploads, network interception — the full Playwright protocol or CDP from Node.js, Python, Java, C# or Go, not a fixed list of scenario steps that must finish in 40 seconds.
- State must last between pages — a session for hours, persistent profiles, a warmup browser for days, or a browser around the clock on a flat-priced dedicated thread.
- The site must see a specific real browser or version — Chrome channels, Edge, Brave, Firefox, WebKit, Camoufox — not an unnamed headless browser.
- You want to watch or take over a session, record it, or share a live tab.
- You want to start free — the Community plan includes proxies — and try a browser in live view before writing any code.
Try it
The quickstart is one HTTPS call and five lines of Playwright; the code builder writes the launch for any browser and option set in your language.