Scavio

Scavio MCP Server

Scavio is a unified Web Search API and MCP server that gives AI agents web search, page extraction, and structured data from e-commerce, social, travel, jobs, real-estate, app-store, ad-library and company-filing sources. 267 tools across 38 platforms plus Extract, map, crawl and batch, one API key.

The 22 platforms added in 0.13.0 are opt-in. Registering all 267 tools puts 254KB of tool definitions — roughly 64k tokens — into every session before you type anything. So the default is the platform set 0.12.x already had, plus Extract: 142 tools, 112KB. Upgrading never removes a tool you were using. Everything else is one env var away: see Choosing which tools load.

Remote MCP Server

Connect directly to Scavio's remote MCP server without any local installation:

https://mcp.scavio.dev/mcp

Two ways to authenticate:

• Sign in with OAuth, no key to copy. Claude Desktop, claude.ai and Claude Code sign in to your Scavio account through the server. In Claude Desktop or claude.ai: Settings > Connectors > Add custom connector, name it Scavio, paste the URL, then Connect and Authorize. In Claude Code: claude mcp add --transport http scavio https://mcp.scavio.dev/mcp, then run /mcp and choose Authenticate.
• API key. Pass it in the x-api-key header. Get your key at scavio.dev.

---

Run Locally (npx)

Prefer to run the server on your own machine? Use npx with no clone or build. The server runs over stdio and only needs your SCAVIO_API_KEY.

Claude Code

claude mcp add scavio -e SCAVIO_API_KEY=YOUR_SCAVIO_API_KEY -- npx -y @scavio/mcp-server

Any MCP-Compatible Client (Claude Desktop, Cursor, Windsurf, VS Code, etc.)

{
  "mcpServers": {
    "scavio": {
      "command": "npx",
      "args": ["-y", "@scavio/mcp-server"],
      "env": {
        "SCAVIO_API_KEY": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

Requires Node.js 20+. Get your API key at scavio.dev.

---

Install

Claude Code

Sign in with OAuth, then run /mcp inside Claude Code and choose Authenticate:

claude mcp add --transport http scavio https://mcp.scavio.dev/mcp

Or pass an API key:

claude mcp add --transport http scavio https://mcp.scavio.dev/mcp --header "x-api-key: YOUR_SCAVIO_API_KEY"

Claude Desktop

No API key needed: Settings > Connectors > Add custom connector, name it Scavio, paste https://mcp.scavio.dev/mcp, then Connect and sign in to your Scavio account. The same steps work on claude.ai.

Or run the server locally. Add to claude_desktop_config.json (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "scavio": {
      "command": "npx",
      "args": ["-y", "@scavio/mcp-server"],
      "env": {
        "SCAVIO_API_KEY": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

Requires Node.js 20+. Restart Claude Desktop after saving.

Cursor

Or add to .cursor/mcp.json in your project root or ~/.cursor/mcp.json for global:

{
  "mcpServers": {
    "scavio": {
      "type": "http",
      "url": "https://mcp.scavio.dev/mcp",
      "headers": {
        "x-api-key": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "scavio": {
      "type": "http",
      "url": "https://mcp.scavio.dev/mcp",
      "headers": {
        "x-api-key": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

VS Code

Create .vscode/mcp.json in your project root:

{
  "servers": {
    "scavio": {
      "type": "http",
      "url": "https://mcp.scavio.dev/mcp",
      "headers": {
        "x-api-key": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

ChatGPT

1. Go to ChatGPT > Settings > Apps & Connectors
2. Click Add custom connector
3. Enter the MCP server URL: https://mcp.scavio.dev/mcp
4. Configure authentication with your API key

Cline

Add to cline_mcp_settings.json (open via Cline sidebar > MCP Servers > Configure):

{
  "mcpServers": {
    "scavio": {
      "type": "http",
      "url": "https://mcp.scavio.dev/mcp",
      "headers": {
        "x-api-key": "YOUR_SCAVIO_API_KEY"
      }
    }
  }
}

Zed

Add to settings (Cmd+,):

{
  "assistant": {
    "mcp_servers": {
      "scavio": {
        "type": "http",
        "url": "https://mcp.scavio.dev/mcp",
        "headers": {
          "x-api-key": "YOUR_SCAVIO_API_KEY"
        }
      }
    }
  }
}

Any MCP-Compatible Client

{
  "type": "http",
  "url": "https://mcp.scavio.dev/mcp",
  "headers": {
    "x-api-key": "YOUR_SCAVIO_API_KEY"
  }
}

---

Choosing which tools load

Scavio exposes 267 tools. Loading all of them costs 254KB of tools/list — roughly 64k tokens of context in every session before the user has said anything, and more tools than some clients will accept at all. So the server registers a curated subset by default and lets you pick the rest. The sizes below are measured, not estimated: npm run toolslist reproduces them.

SCAVIO_PLATFORMS is a comma-separated allowlist of platform keys:

Keys are matched case- and punctuation-insensitively, so meta-ads, metaads and META_ADS are the same thing, and twitter resolves to x. An unknown key is logged to stderr and skipped rather than failing the server.

get_usage always registers — it is free, and it is what every install guide tells you to call to check the server works.

Platform keys: extract, google, chatgpt, gemini, youtube, tiktok, instagram, reddit, x, linkedin, threads, kuaishou, amazon, walmart, ebay, target, homedepot, costco, mercadolibre, tiktok-shop, booking, airbnb, tripadvisor, yelp, zillow, redfin, realtor, rightmove, indeed, glassdoor, appstore, googleplay, g2, capterra, trustpilot, googleads, metaads, sec, companieshouse

Local (stdio)

{
  "mcpServers": {
    "scavio": {
      "command": "npx",
      "args": ["-y", "@scavio/mcp-server"],
      "env": {
        "SCAVIO_API_KEY": "YOUR_SCAVIO_API_KEY",
        "SCAVIO_PLATFORMS": "extract,google,zillow,redfin"
      }
    }
  }
}

Remote (HTTP)

The hosted server has no env var to read, so pass the allowlist per connection — either as a header or as a query parameter:

{
  "scavio": {
    "type": "http",
    "url": "https://mcp.scavio.dev/mcp",
    "headers": {
      "x-api-key": "YOUR_SCAVIO_API_KEY",
      "x-scavio-platforms": "extract,google,sec,companieshouse"
    }
  }
}

https://mcp.scavio.dev/mcp?platforms=all

Upgrading from 0.12.x

Nothing to do. The default set is exactly the 10 platforms 0.12.x registered, plus Extract — every tool you were using still loads. Those platforms keep gaining tools (Walmart grew from 2 to 7; TikTok, YouTube, Reddit and X gained 19 in 0.20.0; Extract gained map, crawl and batch in 0.21.0; Instagram gained 8 in 0.22.0; Google gained google_custom_search in 0.23.0), which is why the default count is now 142.

The 22 platforms added in 0.13.0 are opt-in. Add them additively:

SCAVIO_PLATFORMS=default,zillow,redfin,sec

One breaking fix worth knowing: search_walmart previously sent fulfillment_speed: "anytime" on every call, which the API rejects. The parameter is now today or tomorrow with no default.

---

Available Tools

Every tool costs 1 credit per call unless its section says otherwise. Several platforms price per endpoint (YouTube, Instagram, LinkedIn, Kuaishou) or per request body (Extract, Threads, Walmart); get_amazon_options and get_usage are free.

Extract

Extract is tier-priced by the mode parameter, not flat: normal (the default) and advanced (renders JavaScript) cost 1 credit, ultra (for the hardest pages) costs 2. Start at normal and escalate only when the content comes back empty or blocked. You are billed only on a successful extraction — a dead link, a bot wall or a timeout costs nothing. There is no pagination: the whole page comes back in one call.

Map, crawl and batch (new in 0.21.0) are the multi-page operations:

• Pricing. map_website costs 1 credit. Crawl and batch bill per successful page at Extract's price for the tier (normal 1, advanced 1, ultra 2); failed and skipped pages are free, and status and cancel calls are free. Credits are taken page by page, and a job stops cleanly with stopped_reason: "insufficient_credits" if the balance runs out.
• Plan limits (max pages per crawl / active jobs / URLs per batch): Free 10 / 1 / 10, PAYG 100 / 1 / 50, Project 500 / 2 / 100, Bootstrap 2,000 / 3 / 500, Startup 5,000 / 5 / 1,000, Growth 10,000 / 10 / 1,000. Starting a job past the active-job limit returns a 429 that names the running jobs.
• Behaviour. Robots.txt is always obeyed for map and crawl. A crawl stays on the start URL's site (subdomains optional) and visits pages under the start path first. Results are kept for 3 days; a job runs for at most 6 hours.
• Built for tool-call time limits. MCP clients usually cancel a tool call after about 60 seconds, so a crawl is a job, not one long call. crawl_website / batch_extract start the job and wait up to wait_seconds (default 30, max 50): a small job comes back finished with its first page of results; a large one returns the job id (and webhook_secret, shown only once, if you set webhook_url). get_crawl_status / get_batch_status take the same wait_seconds to long-poll, and cursor to page.
• Output stays small. Status calls return 10 results per call by default (limit up to 100) and clip each page's content to max_content_chars (default 4,000); clipped pages carry content_clipped: true next to the full content_length. map_website asks for 200 links when you set no limit (the API default is 1,000). Every response that has more to read carries a _hint naming the exact next call. Repeat a start call with the same idempotency_key and you get the original job back instead of a second one.

Google Search API

ChatGPT and Gemini

For AI visibility and brand monitoring: ask what a buyer would ask ("best CRM for small agencies") and see whether a brand is recommended, which competitors are, and which pages the answer cites. Both tools return the same shape (answer.markdown, sources[], search_queries[], shopping[]), so comparing assistants needs one parser. Every call is a fresh answer and takes 10-40 seconds. ChatGPT costs 2 credits and Gemini 5 credits per answered prompt; failed prompts are free. Opt-in: add chatgpt and gemini to SCAVIO_PLATFORMS (e.g. default,chatgpt,gemini). Docs: ChatGPT, Gemini.

YouTube Data API

Credit cost varies: get_youtube_transcript costs 5, get_youtube_streams 3, search_youtube and search_youtube_shorts 2, and every other YouTube tool 1. get_youtube_related accepts a cursor but never returns one, so treat it as a single page.

Amazon Product API

Walmart API

Walmart is priced by the request body, not at a flat rate: com (US, default) and ca cost 1 credit, com.mx costs 2. Only search_walmart and get_walmart_category accept every domain.

Store targeting: pass delivery_zip and store_id together on search_walmart or get_walmart_product to get one store's assortment and availability. Take the store_id from get_walmart_stores on the same domain (walmart.com or walmart.ca). A store-targeted call costs 2 credits, takes 10-60 seconds, and echoes the store it used in data.location. get_walmart_product accepts domain: "ca" only together with a store pair.

Two shapes to plan for. get_walmart_offers returns the buy-box winner only — Walmart server-renders just that one offer, so expect a single row even when total_offer_count is higher. And get_walmart_seller_products is capped at roughly the first 40 items with no page or cursor parameter; read total_count to know you are looking at a slice. Both seller tools need the NUMERIC seller_catalog_id from a product, search or offers response — the GUID-form seller_id in those same responses 404s.

TikTok API

Every TikTok tool costs 1 credit. The trending, sound, playlist, similar-creator, live and region tools return TikTok's own payload shape, trimmed for context: large numeric ids can lose precision as JSON numbers, so use the string forms (id_str, sec_uid, sec_user_id). get_tiktok_trending_videos returns about 6 videos per call whatever count says, and get_tiktok_music_videos ignores count (about 14 per page). Search suggestions can be empty for long-tail keywords and still bill. An unknown username is a billed 404 on get_tiktok_user_live and an unbilled 404 on get_tiktok_user_region.

Instagram API

Instagram is priced per endpoint: get_instagram_profile, get_instagram_user_reels, get_instagram_user_tagged and get_instagram_music_posts cost 1 credit (reels and tagged cost 2 the first time a username is looked up), and the other tools cost 2. search_instagram_users and search_instagram_hashtags cost 8 when you pass a cursor for the next page. Any request can cost up to 10 when it needs extra work; credits_used in the response is always the exact charge.

Reddit API

Every Reddit tool costs 1 credit. Two things to know: search_reddit takes a query and a cursor only - there is no sort or post-type filter, results come back in relevance order (search_reddit_comments does sort: relevance, hot, top or new) - and get_reddit_post returns a flat post object with no comments, so call get_reddit_post_comments with the post id for those. Reddit is the slowest platform here, typically 5-15 seconds per call.

X API

get_x_user_tweets, get_x_user_replies and get_x_user_media return a next_cursor but no has_more, so page until the cursor is absent or the timeline comes back empty.

LinkedIn API

All take a vanity handle, slug or id, or a full LinkedIn URL. Credit cost varies: get_linkedin_person, get_linkedin_person_about, get_linkedin_company and get_linkedin_post cost 1; get_linkedin_person_posts, get_linkedin_company_posts, search_linkedin_jobs and get_linkedin_post_comments cost 10 per page; get_linkedin_job costs 30.

Member contact info, the company employee directory, per-company job listings, people search and post search are no longer available, so those five tools were removed. get_linkedin_company still returns a small sample of featured employees, and search_linkedin_jobs with a company name substitutes for per-company listings.

TikTok Shop API

Two things to know: get_tiktok_shop_product does not return a price (TikTok masks it on the product page) - exact prices come from search_tiktok_shop, get_tiktok_shop_shop_products and get_tiktok_shop_category_products. And only about 44% of the product IDs returned by search resolve on get_tiktok_shop_product (11 of 25 measured), because TikTok Shop has no detail data for the rest. That miss is signalled by the HTTP 404 status, not by any field in the response body, and it is a normal outcome rather than an error to retry - the tool returns it as a plain result with status: "no_detail_data" so agents skip the product instead of looping. For an id that will not resolve, get_tiktok_shop_product_reviews is often still usable: across 8 measured ids that failed on detail, 8 of 8 returned HTTP 200 on reviews and 7 of 8 returned at least one review.

Threads

Threads is priced by how you address a user, not per endpoint: 2 credits when you pass user_id, 4 credits when you pass username. Only get_threads_profile, get_threads_user_posts and get_threads_user_replies accept a username; the other three are always 2. Prefer user_id - a handle costs a second lookup.

Kuaishou (China)

Kuaishou is priced per endpoint, not at a flat rate: get_kuaishou_profile and all four search tools cost 10, get_kuaishou_video 2, get_kuaishou_videos_batch 40, and every other tool 1.

eBay

1 credit per call. get_ebay_seller is a profile card and cannot enumerate a catalogue - list a seller's inventory with search_ebay and seller set and no keyword. sold: true searches completed listings that actually sold; on that view eBay publishes no headline count, so total_results is null. per_page accepts only 60, 120 or 240.

Target

1 credit per call. get_target_product, get_target_reviews and get_target_availability are single-shot; search and category page with page + count. store_id on search, category, product and reviews sets the store for prices only; get_target_availability is the store-stock tool.

Home Depot

2 credits per call. Search page size is fixed at 12 and cannot be changed. Reviews return 30 per page; total_pages is the last page that exists and asking past it is a 404.

Costco

1 credit per call, except the multi-warehouse tools: get_costco_prices makes one lookup per two locations (online counts as one) and get_costco_availability one per warehouse, billed 1 credit per 5 lookups, so 10 warehouses cost 2 credits. get_costco_warehouses and get_costco_gas_prices cost 2 for country: "ca"; gas by warehouse_ids is 1 credit per 5 warehouses (2 per warehouse on ca).

search_costco, get_costco_category and get_costco_deals take warehouse_id to add that warehouse's in-store price, stock and price code to every result. Get the id from get_costco_warehouses. country is us (default) or ca everywhere except get_costco_coupons (US only); search_costco, get_costco_product and get_costco_warehouses also accept uk, au, mx, jp, kr and tw for keyword, sort and paging only. Price codes in get_costco_clearance are the member-community decode (.97 clearance, .00/.88 manager markdown; .49/.79/.89 special buy on request), not something Costco publishes, and only items Costco lists online are covered. get_costco_coupons returns the first 50 offers of a ~200-offer book by default; pass limit for more.

Mercado Libre

Mercado Libre (and Mercado Livre in Brazil) product data for Mexico, Brazil, Argentina and Colombia: listings, prices, sellers, reviews, buyer questions and category best sellers, with no OAuth app to register.

2 credits per call; get_mercadolibre_reviews is 1. Search serves up to 42 pages (about 2,000 results) per query, so slice by price range, condition or seller to reach more, and sorts by relevance or price only. Units sold are lower bounds ("at least 500"), not exact counts. Seller name appears on a search row only where Mercado Libre shows it; get_mercadolibre_product returns the buy-box seller, and the list of every seller on a product is not available. Mercado Libre may narrow a query on its own (Brazil and Argentina can apply "new"); auto_applied_filters lists what it applied. Product, reviews and questions take no country: the id prefix (MLM, MLB, MLA, MCO) or URL sets it. Chile is not supported. A zero-result search is an empty, billed page; an unknown product, item or seller nickname, or a listing with no questions, is a billed 404 that says so.

Booking.com

1 credit per call. Search returns 25 properties per page; get_booking_hotel and get_booking_reviews are single-shot.

Airbnb

1 credit per call. search_airbnb takes page XOR cursor - sending both is rejected, and cursor wins. 18 listings per page. get_airbnb_reviews pages with limit + offset.

Tripadvisor

2 credits per call. Start with resolve_tripadvisor_location for the geo_id / location_id pair everything else needs. Search returns 30 per page and a page past the last is a 404, not an empty result. Reviews return 15 per page for restaurants and 10 for hotels and attractions, so category must match the location's own type on any page past the first.

Yelp

2 credits per call. Yelp fixes the page size at 10. get_yelp_business already includes the first page of reviews at no extra cost. A review page past the last is a 404, not an empty result.

Zillow

1 credit per call. Only search_zillow paginates. get_zillow_agent_reviews is an AGENT profile, not a property.

Redfin

1 credit per call. search_redfin pages with page + limit, up to 350 listings per page; the other two are single-shot.

Realtor.com

1 credit per call. search_realtor returns up to 200 listings per page (default 50) and serves the first 10,000 results of a search (page x page_size <= 10,000); slice a big area by price band, property type or radius to reach more. Sold prices are null in US non-disclosure states (Texas, Utah, Wyoming and about nine others), while sold dates still come back. An unknown property or agent id is a billed 404.

Rightmove

1 credit per call. Rightmove search has no keyword filter: match on the returned summary and key_features instead. Paging caps per search: 42 pages for search_rightmove (about 1,000 listings), 40 for get_rightmove_sold, 50 for search_rightmove_agents; narrow by price band, type or a smaller area to reach more. Sold prices cover England and Wales only (not Scotland or Northern Ireland). Live listings carry no price history, so use get_rightmove_sold_property for an address's past sales. A removed listing or an unknown id is a billed 404.

Indeed

2 credits per call. Job search returns 10 postings per page, company reviews 20 per page. get_indeed_job and get_indeed_company are single-shot.

Glassdoor

1 credit per call. Start with resolve_glassdoor_company for the employer_id. Only get_glassdoor_salaries paginates (10 job titles per page; page_count is how many exist). get_glassdoor_reviews returns up to three full reviews plus complete rating statistics - it is not a paginated review feed.

Apple App Store

1 credit per call. Reviews page 1-10 at 50 per page and stop hard at page 10: 500 reviews per storefront is Apple's anonymous ceiling, and you reach further by asking a different country. Search and app detail do not paginate.

Google Play

2 credits per call - a separate namespace from google and not on the Google-exempt price. Reviews are cursor-paginated, and the cursor is opaque, single-use and encodes the sort as well as the position, so send it back with the same sort it came from. A cursor past the last review is a 404. Search and app detail do not paginate.

G2 Software Reviews

5 credits per call - the most expensive platform on Scavio. search_g2_software pages with page + limit (capped at 100 on our side); get_g2_reviews is fixed at 10 per page and paginates well past the 10 pages G2's own widget links to. get_g2_product is single-shot.

Capterra Software Reviews

2 credits per call. Only get_capterra_reviews paginates: 25 per page, capped at page 100 - past it Capterra answers 200 with page one, so check what came back. get_capterra_product already includes the 25 most recent reviews.

Trustpilot

2 credits per call. get_trustpilot_reviews serves 10 pages of 20, so 200 reviews per filter combination; to read more, slice into separate sets (each stars value, each language, each date_range) rather than paging further. total_filtered with a date_range doubles as review velocity. Email and phone appear only when the business published them; the AI summary and topics only when Trustpilot has them for the chosen language (default en). No TrustScore history. An unknown domain, category or review id is a billed 404 that names what was not found.

Google Ads Transparency

1 credit per call. Start with resolve_google_ads_advertiser to turn a brand name or domain into an advertiser_id. search_google_ads is cursor-paginated at 100 ads per page - re-send the same filters alongside the cursor. get_google_ads_creative is the only tool carrying a creative's history, region breakdown and impression bucket.

Meta Ad Library

1 credit per call, and every cursor page is another credit: page 1 returns 30 ads, then 10 per page thereafter, so the cost of walking a whole query scales with depth. search_meta_ads and get_meta_ads_advertiser both cursor-paginate - walk has_next_page. get_meta_ad is single-shot.

SEC EDGAR

1 credit per call on the official free EDGAR JSON API. Start with resolve_sec_company to turn a name or ticker into a CIK. get_sec_filings pages with page + limit; search_sec_filings pages up to 100, because the index refuses a result window past 10,000.

Companies House

1 credit per call on the official UK register. Start with search_companies_house for a company_number. Search returns 20 per page and is capped at page 50 - the register serves a 1000-result window per term whatever hit count it prints, then answers page 51 with a 416. Officers (35/page) and filing history have no upper page bound; past the last page they return an ordinary 200 with an empty list, which looks identical to having none.

Account

---

About Scavio

Scavio is a unified search API for AI agents and a data API for developers. One key, structured JSON:

• Google Search API — SERP results, news, images, maps, and knowledge graph
• Extract — read any URL as Markdown, plain text or raw HTML
• Amazon Product API and Walmart Product API — product search and details, alongside eBay, Target, Home Depot and Costco (per-warehouse prices, stock, gas prices, coupon book), and Mercado Libre (Mexico, Brazil, Argentina and Colombia listings, sellers, reviews, questions and best sellers)
• TikTok API, Instagram API, Reddit API, X API, LinkedIn API, and YouTube API — social and video data
• Travel and local — Booking.com, Airbnb, Tripadvisor, Yelp
• Real estate — Zillow, Redfin, Realtor.com (listings, sold comps, market stats, agents), Rightmove (UK listings, England and Wales sold prices since 1995, agents)
• Jobs and employer data — Indeed, Glassdoor
• Apps, software reviews and ad libraries — App Store, Google Play, G2, Capterra, Trustpilot (business profiles, filtered reviews, category rankings), Google Ads Transparency, Meta Ad Library
• Company filings — SEC EDGAR, Companies House

Teams use it as a SerpAPI alternative with structured multi-platform data — see Tavily vs Scavio for a head-to-head comparison.

Get a free Search API key at scavio.dev.

---

License

MIT