Skip to content
ScrapingBot logo

ScrapingBot MCP server

Which servers in the official MCP registry are ScrapingBot's own, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-21

← Back to ScrapingBot

Whose server it is
  • Registry nameio.scrapingbot/scrapingbotverified in the MCP registry

The registry name sits under scrapingbot.io. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-21
  • Stateanswerschecked 2026-10-21
  • Tools23
  • Server calls itselfscrapingbot-mcp
  • Server version1.0.0
  • Protocol version2025-06-18

On 2026-10-21 the server answered initialize and tools/list without a sign-in. First answer in 322 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: fetch rendered pages that need JavaScript or proxies; collect TikTok video stats and Instagram post data; pull Google search results and Amazon product rankings.
  • Checked on 6 of the 7 days it takes to judge reliability.

Once connected

  • For "fetch rendered pages that need JavaScript or proxies", the likely first call is scrapeWebsite.
  • For "pull Google search results and Amazon product rankings", the likely first call is googleSearch.
  • Parameters are untyped or undescribed in 21 of 23 tools: check arguments before calling.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 6, the first on 2026-10-21
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 6 so far
  • What an agent can do with it3 of 3 use cases
  • Answeredmeasuring6 of the 7 daily checks it takes to print a share
  • Median first answer288 ms
  • Space in an agent's context≈ 3,876 tokensthe definitions of all its tools (23), which an agent carries on every turn
  • Tool definitionsB · 85 of 100
  • Typed results0 of 23 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 21 of 23

What an agent can do with it

  • can do · fetch rendered pages that need JavaScript or proxies
  • can do · collect TikTok video stats and Instagram post data
  • can do · pull Google search results and Amazon product rankings

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (30%), doing the product's job (25%), definitions (20%), context (10%) and use (15%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers23 tools · read 2026-10-21
  • listCapabilities

    List every ScrapingBot tool, the API routes behind them and their parameters. Free.

  • scrapeWebsite

    Fetch any public web page and return its content as markdown (format: "html" for the raw HTML). 1 credit; render_js (pages built with JavaScript) 5; premium_proxy 10; stealth_proxy 75. screenshot: true attaches a screenshot image.

  • extractStructuredData

    Fetch a web page and have AI pull out exactly the fields you want, as JSON. Pass ai_query (plain English, e.g. "title, price and stock") or ai_schema ({"field": "description"}). The page cost plus 5 credits.

  • getScrapeJob

    Look up an earlier scrape by the job_id it returned. Free.

  • pollJobUntilDone

    Wait for a scrape job to finish, then return it. Free.

  • runBrowserScenario

    Open a page in a real browser, run steps such as click, fill, scroll and wait, then return the resulting page as markdown. js_scenario: {"instructions": [{"fill": ["#q", "shoes"]}, {"click": "#go"}, {"wait": 1000}]}. From 5 credits.

  • googleSearch

    Google results as JSON: web (default), images, videos, news, shopping, places or maps (type), for any country (gl) and language (hl). Places and maps results include a cid for googleReviews. 10 credits.

  • googleReviews

    Reviews of a place from Google Maps, by the cid, fid or placeId in a googleSearch places or maps result. sortBy newest, mostRelevant, highestRating or lowestRating; pass nextPageToken for more. 10 credits.

  • instagramUser

    An Instagram profile: bio, follower, following and post counts, links and the numeric user id other tools need. Pass one of username, user_id or url. 5 credits.

  • instagramSearch

    Search Instagram accounts, hashtags, places or posts (search_type), or everything at once (global). 5 credits.

  • instagramMedia

    Instagram content by mode: an account's posts, reels or tagged posts (user_id), one post by shortcode or url, a post's comments, the replies to a comment (comment_id), or an account's active stories (username). 5 credits.

  • instagramFollowers

    List or search an Instagram account's followers or following, by numeric user_id from instagramUser. query searches the whole list; max_id pages. For very large accounts Instagram shows only about 50 recent followers, so use query to find…

  • tiktokVideo

    A TikTok video's stats, caption, author, sound and download links, by video URL. 1 credit.

  • tiktokUser

    A TikTok profile by username: bio, counts, verified badge and numeric user id; include_posts returns the latest videos instead. 1 credit.

  • tiktokSearch

    Search TikTok creators (search_type: users) or videos (videos) by keyword. Videos page: for page 2 onward pass the previous response's data.cursor as cursor and its data.searchId as search_id. Users search returns a single page and…

  • tiktokComments

    Comments on a TikTok video (url), or the replies to one comment (comment_id plus the video url or video_id). 1 credit.

  • tiktokFollowers

    A TikTok account's followers or the accounts it follows, by numeric user_id (data.user.id from tiktokUser). Up to 50. 1 credit.

  • amazonSearch

    Search Amazon US products by keyword: title, price, rating, Prime and photo for each. page for more. 10 credits.

  • amazonProduct

    Everything on an Amazon product page by ASIN: price, rating, availability, bullet points, specifications, photos and variations. 10 credits.

  • amazonProducts

    Full details for up to 20 ASINs in one call, the same as amazonProduct for each. 10 credits per product returned; ASINs that fail are free.

  • amazonSuggestions

    The suggestions Amazon shows as a shopper types: useful for keyword research. 10 credits.

  • amazonRankings

    Amazon's Best Sellers or New Releases for a department, in rank order, with price and rating. 10 credits.

  • providerRequest

    Call any Google, Instagram, TikTok or Amazon endpoint by path, for the few the other tools do not cover (listCapabilities lists them). Costs the same as that API.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "scrapingbot": {
      "type": "http",
      "url": "https://scrapingbot.io/api/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http scrapingbot https://scrapingbot.io/api/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://scrapingbot.io/api/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.