Scraping API pricing

This skill answers plan and price questions about web data APIs with the public Web Data API Plan Finder service (https://plans.intoperson.com). It needs no key and no account. Every request below is a plain HTTPS call.

When to use

• The user needs a scraping, crawling, SERP, proxy or headless browser API and asks which plan fits a monthly volume, a budget or a billing rule.
• The user names plans and asks whether they meet a condition.
• The user asks how the cost changes when the workload grows.

Do not answer these from memory. Prices and terms change, and the service records each fact with its source sentence and the date it was checked.

Steps

1. Turn the request into conditions. Ask only for what changes the answer: pages per month by type (plain_html, javascript, premium_proxy, javascript_premium_proxy), or monthly_pages_total if the split is not known; monthly_searches for SERP; max_monthly_cost_usd; and any must-have rule such as failed_requests_not_billed, search_engines, search_countries (ISO alpha-2, e.g. "DE") or search_result_types.

2. Search:

   curl -s -X POST https://plans.intoperson.com/api/v1/search \
     -H "Content-Type: application/json" \
     -d '{"monthly_searches":20000,"search_engines":["google"],"search_countries":["US"],"max_monthly_cost_usd":100}'
   

   curl -s -X POST https://plans.intoperson.com/api/v1/search \
     -H "Content-Type: application/json" \
     -d '{"monthly_pages":{"javascript":50000},"max_monthly_cost_usd":200}'
   

   On Windows PowerShell call curl.exe, not curl.

3. Read the result:

   • full_matches meet every required condition. partial_matches have conditions listed in unmet_conditions or unknown_conditions.
   • monthly_cost.explanation shows how the cost was computed. A cost can be a range (low, high) when the price depends on the target site.
   • Each fact has evidence (official page URL and quoted sentence) and checked_at.

4. If nothing fits, ask which conditions can change and call /api/v1/relax with the same body plus "negotiable": [...] (for example "monthly_budget", "billing_type", "serp_engines"). It returns the smallest changes that admit a plan.

5. Other calls, all with the same condition fields:

   • POST /api/v1/check with "offer_ids": [...] or "plans": ["ScrapingBee Freelance"] checks named plans.
   • POST /api/v1/scenarios with "scale": 2 or a "target" workload shows plan costs as the workload grows.
   • GET /api/v1/offers/{offer_id} and GET /api/v1/products/{product_id} return every recorded fact of a plan or product.
   • Full schema: https://plans.intoperson.com/api/v1/openapi.json

How to report

• Lead with the plans that meet every condition, their monthly cost and the cost explanation.
• For each claim you pass on, give the quoted sentence, the page URL and the check date from the response.
• When a condition is UNKNOWN, say the official page does not state it. Do not fill it in from general knowledge.
• For a follow-up question about one product (for example "do unused credits roll over?"), read its terms with GET /api/v1/products/{product_id} and answer from the matching fact. The service records prices, included units, overage, rollover, concurrency, rate limits, timeouts, rendering, proxies, geo targeting, SERP engines, countries and result types, crawl limits, failed-request billing, free allowances, tax and commitments. For anything else (refunds, SLA, terms of use, data retention), say the service does not record it and point to the provider's page.
• Say that prices exclude tax and that the user should confirm the final price and account limits with the seller before buying. Order is by monthly cost.

MCP server (optional)

The same service is an MCP server with the tools search, check_conditions, cost_scenarios, relax_conditions and get_details:

claude mcp add --transport http web-data-plan-finder https://plans.intoperson.com/mcp

Other clients: add the streamable HTTP URL https://plans.intoperson.com/mcp.