Web Search MCP Servers: Tools, Connections & Account Requirements

• search-actors

  Find Actors in Apify Store.[2]

• fetch-actor-details

  Read an Actor input schema, description and pricing.[2]

• call-actor

  Start the selected Actor with its documented input.[2]

• get-actor-run

  Read a run status and its associated storage identifiers.[2]

• get-dataset-items

  Read dataset rows with pagination and filtering.[2]

The configured token uses the Apify account balance. Actor pricing can be per event or platform usage; compute, proxies, storage and data transfer may contribute. Inspect the selected Actor pricing before execution.

Required product: Apify Actors and platform usage

• search_engine

  Retrieve results from a supported search engine.[3]

• scrape_as_markdown

  Read one page as Markdown.[3]

• scraping_browser_navigate

  Navigate a remote browser session (browser group; enabled by default unless groups/tools restrict the server).[4][3]

API usage draws from the account’s allowances; additional usage and optional tools follow the applicable product terms. Free server software does not make all API usage free.

Required product: Bright Data Web Unlocker API; additional products may be used by optional tools.

• scrape_as_markdown

  Return a requested page as Markdown.[1]

• google_search

  Retrieve parsed Google results.[1]

• amazon_product

  Retrieve a supported Amazon product page.[1]

Scraping usage is billed under the account plan; request features such as rendering and proxy pool affect consumption.

Required product: Decodo Web Data API.

• list_proxy_products

  Inspect account products, gateways, ports and balances.[1]

• list_proxy_targeting_options

  Find supported locations and network targeting values.[1]

• get_proxy_usage

  Read recent bandwidth consumption.[2]

• rotate_proxy_session

  Change the exit IP of a sticky session.[2]

• scrape_url

  Retrieve a page with scraping options, including documented browser mode.[1]

• search_domains

  Find domains from a search query.[2]

Scraping consumes credits according to mode and options. Proxy traffic uses the applicable product allowance. The software license does not include API usage.

Required product: Evomi Public API access plus the proxy or Scraper API product used by the tool.

For this task: The default hosted tools search and fetch known pages. Agent requires authentication. An explicit tools parameter replaces defaults, so include every operation needed by the workflow.[1]

• web_search_exa

  Search the web for source pages and relevant content; enabled in the default hosted tool set.[1]

• web_fetch_exa

  Fetch content from known URLs; enabled in the default hosted tool set.[1][4]

• web_search_advanced_exa

  Opt-in search with additional controls and filters. An explicit tools list replaces defaults.[1]

• agent_run

  Authenticated hosted research. A running response returns an id that can be passed back as runId to wait for the same work.[1]

Account usage follows the operation: search requests, returned content pages and Agent effort have different prices. Credits are shared across a team’s keys, with optional per-key budgets. Consult the current pricing table before enabling research.

Required product: Rate-limited hosted search/fetch without an account. OAuth or a team API key enables account usage and the hosted Agent tool.

For this task: Hosted keyless access exposes Search, Scrape and Parse. Choose OAuth or a bearer key for the account tool surface; anonymous daily caps are separate from the team plan.[2][5][1]

• firecrawl_search

  Find web results from a query; available in keyless mode.[4]

• firecrawl_scrape

  Read a known URL; JSON format can extract requested fields. Available in keyless mode.[4]

• firecrawl_parse

  Parse supported files. Hosted local-file input requires an upload handoff.[4]

• firecrawl_map

  Discover site URLs with authenticated access.[4]

• firecrawl_crawl

  Traverse a site scope with authenticated access; normally waits for a terminal job state.[4]

• firecrawl_check_crawl_status

  Resume a crawl wait using its job ID.[4]

• firecrawl_interact

  Operate a live page with authenticated access.[4]

• firecrawl_interact_stop

  Release an Interact session using its scrapeId.[4]

Keyless use has daily per-IP caps. Account use consumes plan credits and any option charges; returned target error pages can be billable. Scrape JSON adds extraction credits, and browser sessions use separate time-based billing.

Required product: Keyless hosted Search, Scrape and Parse; a Firecrawl account/team for authenticated tools.

• list_runs

  Read the account’s runs and apply documented filters.[1]

• create_run

  Start a data collection run; requires runs:write.[1][2]

• get_run

  Read a run and its item statuses.[1]

• export_run

  Export collected results in a supported file format.[1]

• compare_runs

  Create an analysis comparing collected runs; requires analyses:write.[1]

• create_schedule

  Create a recurring run schedule; requires schedules:write.[1]

• get_usage_summary

  Read the account’s usage summary with daily detail.[1]

Platform access is included with eligible Web Scraper Pro plans and uses the same request quota as the API. Check the current plan eligibility and account quota before creating runs.

Required product: Infatica Data Platform / Web Scraper Pro

• nimble_search

  Find web sources with configurable search behavior.[1][2]

• nimble_extract

  Retrieve content from a known URL.[1][2]

• nimble_map

  Discover website URLs using links and sitemaps.[1][2]

The published API model charges for successful requests, with different rates by product. Search depth and requested content can affect search cost. Managed Data Services plans and research-agent task pricing are separate considerations.

Required product: Nimble API usage

• novada_setup

  Check API-key configuration and account readiness.[3]

• novada_discover

  Inspect the tools available on this connection.[2]

• novada_search

  Find pages using a search query and supported filters.[2]

• novada_extract

  Retrieve content from selected URLs.[2]

• novada_browser

  Perform documented page interactions (local stdio only; excluded from the hosted mcp.novada.com surface).[2]

• novada_proxy

  Produce proxy connection settings for an HTTP client.[2]

Scraping products draw from shared credits, with consumption depending on the product and request. Proxy plans and other product quotas can impose separate limits. Check both balance and the selected product.

Required product: Novada account with access to the scraping, browser or proxy product selected by the tool.

• universal_scraper

  Retrieve page content with optional rendering.[1]

• google_search_scraper

  Retrieve Google results.[1]

• ai_browser_agent

  Run a prompted browser task with AI Studio credentials.[1]

Web Scraper API plans bill by successful results, with target and rendering affecting consumption. AI Studio has separate product usage terms.

Required product: Oxylabs Web Scraper API for the included setup; AI Studio requires its own key.

• scrape

  Fetch a URL with optional rendering and output formatting.[2]

• google_search

  Return parsed Google Search results.[2][1]

• google_shopping

  Return parsed Google Shopping product results.[2][1]

Successful requests consume credits from the account plan. Target sites, structured endpoints and options such as rendering or premium routing affect request cost. Plans also have concurrency limits.

Required product: ScraperAPI API credits

• fast_search

  Retrieve structured web search results.[2]

• get_page_text

  Retrieve a page as Markdown or text.[2]

• extract_page_data

  Extract fields using supported selectors or extraction options.[2]

Calls consume API credits according to the chosen features and proxy options; account plan limits apply.

Required product: ScrapingBee API account.

• my_proxy_status

  Read the current exit, carrier and remaining quota.[1]

• fetch_url

  Fetch a URL as Markdown, text or HTML through the provider’s Singapore proxy.[1]

• search_google

  Retrieve Singapore-localized Google organic results.[1]

• rotate_ip

  Request a new Singapore mobile exit IP.[1]

MCP onboarding offers a limited trial; continued use requires the provider’s paid access. Confirm remaining quota and plan terms in the account.

Required product: Singapore Mobile Proxy MCP access and its proxy plan.

For this task: Send the explicit keyless header for Search and Extract. Account usage can use OAuth or a bearer key; OAuth may prefer a personal key over a team key. Search depth and successful extracted URLs have different credit units.[2][1][4]

• tavily_search

  Find web results for a query. Basic and advanced search use different credit amounts on account plans.[2][4][6]

• tavily_extract

  Retrieve content from provided URLs. Extraction depth affects account credit usage; successful URLs are the billing unit.[2][4][6]

Account basic search costs 1 credit; advanced search costs 2. Basic extraction costs 1 credit per 5 successful URL extractions, advanced 2 per 5. Failed extraction is not charged. These rules do not define anonymous caps or a universal failed-request refund.

Required product: Keyless hosted Search and Extract for light use. An account key, supplied directly or selected through OAuth, is needed for broader operations.

• list_engines

  List engines and their schema resources.[2]

• search

  Request search results using an engine-specific parameter object.[2]

• history

  Read SERP request history.[2]

• statistics

  Read SERP usage statistics for a date interval.[2]

SERP plans provide response allowances. The billing guide charges successful HTTP 200 searches, including successful responses with few or no matching results; other status codes are not billed.

Required product: Thordata SERP API.