Scrapling
D4Vinci/ScraplingReviewed by AI
Checked by RepoPicks Radar on
RepoPicks Radar found strong risk signals in this project. It stays visible so you can read the evidence, but it is left out of recommendations and only appears in the rankings of the “With alerts” filter.
What is Scrapling?
A Python library and terminal command for collecting data from websites. It can make plain web requests, drive a hidden browser for pages that need JavaScript, follow links across a whole site with several pages at a time, pause and resume, and save the results as JSON, JSONL, CSV, XML or clean Markdown. It also ships a server so AI assistants can browse and pull data through it. It is made for programmers, although the `scrapling extract` command saves a page without writing code. You need Python 3.10+ and, for the browser part, a one-time download of a browser.
How easy is it to install Scrapling?
C · install 1 or 2 tools · Windows
- Windows C
- macOS C
- Linux C
The same on Windows, macOS and Linux: install Python 3.10+, then `pip install "scrapling[fetchers]"` and `scrapling install` to download the browser, and use it from a terminal or from your own Python code. A ready Docker image with the browsers inside exists as well, which is one tool more to install.
Scrapling on GitHub
- Stars
- 101,070on Nov 10, 2026
- Downloads
- —all releases
- Growth
- +6,278stars in 7 days
Stars over time
Facts
- Runs on
- Windows, macOS, Linux
- You need to install
- Python, Docker
- To build from source
- Python, Docker
- Disk space
- 100 MB–300 MB · estimate
- Accounts or keys
- None
- Cost
- Free · The library is free (BSD-3-Clause). Large jobs usually mean paying a proxy company, and the README is sponsored by several of them.
- License
- BSD-3-Clause (Permissive)
- Can replace
- Browserbase, Bright Data
- Last change on GitHub
- Nov 9, 2026
Scrapling risk signals (7)
- Highgets around website bot protection
O README lista como função principal: "Anti-bot Bypass: Advanced stealth capabilities with `StealthyFetcher` and fingerprint spoofing. Can easily bypass all types of Cloudflare's Turnstile/Interstitial with automation"; os exemplos usam `solve_cloudflare=True` e a seção do servidor MCP diz que ele usa "our stealthy browser with the right tools to enable suitable AI models to bypass any CAPTCHA is faces while working (not to mention our Cloudflare solver)". O `pyproject.toml` descreve o projeto como "undetectable" e repete a palavra entre as keywords. Passar pelos sistemas que um site usa para bloquear acesso automático vai contra os termos de uso da maioria dos sites, e vários países têm leis sobre contornar proteções técnicas de acesso. O próprio aviso do README diz que a biblioteca é "provided for educational and research purposes only" e pede "Always respect the terms of service of websites and robots.txt files", enquanto a lista de spiders mostra que o respeito ao robots.txt é um "Optional `robots_txt_obey` flag".
AI review
- Attentionbuilt to get around website bot protection
The README sells evasion as a main feature: "Anti-bot Bypass: Advanced stealth capabilities with `StealthyFetcher` and fingerprint spoofing. Can easily bypass all types of Cloudflare's Turnstile/Interstitial with automation", code samples use `solve_cloudflare=True`, the MCP section says it lets AI models "bypass any CAPTCHA is faces", and `pyproject.toml` describes the project as "undetectable". Getting past the systems a site uses to block automated access goes against the terms of use of most sites, and in some countries it can also break the law. The README's own caution says the library is "for educational and research purposes only" and asks you to "respect the terms of service of websites and robots.txt files".
AI review
- Attentionrobots txt is off by default
README Spiders list: "Robots.txt Compliance: Optional `robots_txt_obey` flag that respects `Disallow`, `Crawl-delay`, and `Request-rate`". The rules a site publishes for crawlers are only followed if you switch that flag on.
AI review
- Infocan copy whole catalogues and hide your address
Ready-made templates include `ShopifySpider` to "collect products from public Shopify stores, one item per variant", and the fetchers offer proxy rotation, TLS fingerprint faking (`impersonate='chrome'`) and "DNS Leak Prevention" through Cloudflare DNS-over-HTTPS. Collecting personal data this way also brings privacy-law duties.
AI review
- Inforeadme is paid advertising space
A large part of the README is "Platinum Sponsors" and "Sponsors" blocks for proxy and CAPTCHA-solving companies with discount codes ("Use code SCRAPLING35 for 35% discount", "Code SCRAPEANDTAKE: 25% off") and the line "Do you want to show your ad here?". Treat the recommendations there as ads, not advice.
AI review
- Infoai server feeds outside web pages to your assistant
The MCP server "Let[s] AI chatbots and agents (Claude/Cursor/etc) scrape through Scrapling with HTTP requests, browser actions, screenshots, saved login state". The README says it "protect[s] against Prompt-Injection", which is the owner's claim; text on a page it visits still reaches your assistant.
AI review
- Infosaved logins and captured traffic land on disk
The browser sessions keep cookies and state (`StealthySession`, "saved login state") and `record_requests=True` "saves text response bodies" for later searching. Those files can hold your logins and the data of the sites you visited.
AI review
Evidence is shown in its original language.
What Radar could not check about Scrapling
- Whether the Cloudflare and CAPTCHA handling still works today, since sites change their protection constantly
- The speed numbers in the README, which were measured by the project itself
- Real disk use after `scrapling install` downloads a full browser and its system files
- Whether a crawl stays polite enough not to get your address blocked by the target site
- Whether the claim of protection against prompt injection in the AI server holds up