MCP servers for Amazon, SERP and web scraping
FastAF is one option for agents that need Amazon data, Google results and individual web pages through MCP. We build FastAF, so this is a selection guide rather than an independent ranking. Choose by your workload and test representative requests before committing.
How should you judge a server?
Compare hosting responsibility, authentication, output shape and billing. Decide whether you need individual pages, a recursive crawl, current product snapshots or historical observations. Ask how failures are billed and which parameters the MCP tool actually exposes; a REST option is not automatically an MCP option.
Which alternatives should you evaluate?
Firecrawl, Apify, Playwright-based MCP integrations, SerpApi, Brave Search, Keepa and Amazon-focused APIs such as Rainforest are names to investigate. This guide makes no claims about their current tool counts, prices, speed, coverage or required accounts. Verify those details in each provider’s current documentation, then compare the same workload. A browser-oriented integration, a search service and an Amazon-data service need not solve the same problem.
What does FastAF expose?
| Task | MCP tool | Separate x402 REST price |
|---|---|---|
| Google organic results | google_search |
$0.003 for /serp |
| Amazon search or product data | amazon_product |
$0.003 for /amazon |
| One page as Markdown | scrape_url |
$0.005 for /scrape |
Prices checked against code and live 402 quotes on 27 October 2026. These are wallet-paid REST prices, not MCP prices. MCP uses account credits, including the 500 free monthly credits. Connect to https://api.fastaf.co/mcp using Streamable HTTP. Discovery is public; tool calls need an API key or OAuth token authorized with an existing FastAF key. There is no x402 payment on this MCP URL.
google_search returns Google results with country and language inputs. News, Shopping, Scholar, Jobs, Maps and AI answers have separate tools. amazon_product supports product lookup and keyword search. For offers and bestsellers, use the REST Amazon endpoint with resource=offers or resource=bestsellers. Store your own dated snapshots for history; FastAF does not persist scraped payloads.
scrape_url returns Markdown and exposes render_js, premium, challenge_mode and max_chars. It does not expose all REST output formats or REST’s wait parameter. Its Markdown cap defaults to 12,000 characters, configurable from 1,000 to 30,000. REST scrape supports one selected format: HTML, Markdown, links, images or summary. AI summary adds credits and is outside the current fixed x402 scrape price.
When does FastAF fit?
Use it when your workflow needs structured Google or Amazon results alongside individual page extraction. Scrape accepts one URL per request, not a recursive crawl. Test your target and required fields, retain source URLs and timestamps, and budget each scheduled observation. The OpenAPI reference describes the wallet-paid HTTP surface.
FAQ
Does the hosted MCP URL accept wallet payments?
No. MCP calls use API-key or OAuth authentication. The separate x402 REST routes accept wallet payments.
Does FastAF store Amazon price history?
No. Store your own snapshots with timestamps, marketplace and currency if you need a historical series.
How can I test without purchasing credits?
Register and verify your email for 500 free credits every month per account, shared across keys and REST/MCP. Public MCP discovery does not consume credits.
Where can I explore the API?
Read the documentation and pricing, or try the browser tools. Relevant products include the Google SERP API, Amazon API and web scraping API. See Amazon price monitoring, page-to-Markdown workflows and the x402 glossary.
Reference
- FastAF documentation: every parameter and response field
- Docs: authentication, errors, MCP and x402
- Pricing: credit packs from $10, without expiry
Other sources
- How to read public Amazon data when your requests fail
- How to read public Google Search data when your requests fail
- How to read public Walmart data when your requests fail
- How to read LinkedIn data with an API
- How to read public eBay data when your requests fail
- How to read public Booking.com data when your requests fail
- How to read X profile data with an API
- How to read public Zillow data when your requests fail
- Pay-per-call web scraping for AI agents with x402
- Cost per agent: monthly SaaS vs x402 pay-per-call
- How FastAF handles heavy pages and timeouts
- x402 endpoint discovery checklist: how agents find a paid API (and how FastAF passes it)
- Add FastAF to Claude Code, Codex and Cursor
- How to detect poisoned or incomplete SERP results in AI agents
- Cap agent spend per call with fixed x402 prices
- Use FastAF from the Cloudflare Agents SDK
- FastAF as a fallback search provider
- Budget a scraping job before you run it
- Try FastAF for free, no card
- Too many MCP tools? Keep agent context small
- How agents find an API (and where FastAF is)
- Why agent retries cost more than you think
- All how-to guides