# ScrapeSelect

> Current pricing, normalized monthly costs and recommendations for web scraping APIs, web unlockers and browser infrastructure. Every number cites an official provider page with the date it was checked. Estimates are not invoices; the provider's checkout is authoritative.

## Tools for agents

- [MCP server](https://scrapeselect.com/mcp): Streamable HTTP. Tools: search_providers, get_provider, compare_products, estimate_cost, recommend_scraping_stack, get_recent_changes
- [MCP Server Card](https://scrapeselect.com/mcp/server-card): the MCP server's identity and endpoint, also listed in [/.well-known/ai-catalog.json](https://scrapeselect.com/.well-known/ai-catalog.json) and the [MCP Registry](https://registry.modelcontextprotocol.io/v0/servers?search=com.scrapeselect) as com.scrapeselect/scrapeselect
- [OpenAPI](https://scrapeselect.com/openapi.json): the HTTP API
- POST https://scrapeselect.com/api/v1/recommend: architecture + top 3 providers for a workload, with fit scores, cost ranges, reasons, limitations and evidence
- POST https://scrapeselect.com/api/v1/estimate: monthly cost range per product and the cheapest plan that fits
- POST https://scrapeselect.com/api/v1/compare: up to three products side by side

## Data

- [All providers](https://scrapeselect.com/api/v1/providers): JSON
- [Pricing changes](https://scrapeselect.com/changes.txt): changes confirmed against official pages ([JSON](https://scrapeselect.com/changes.json))
- Every provider and product page has .json and .txt twins, e.g. https://scrapeselect.com/providers/zyte.json

## Guides

- [What 500,000 pages a month costs on 8 scraping APIs](https://scrapeselect.com/guides/scraping-api-cost-500k-pages) (2026-09-26): One workload priced on nine products at three difficulty levels. The cheapest option for easy pages is not the cheapest for hard ones, and the billing unit explains why.

## Products

- [Apify Apify Platform](https://scrapeselect.com/products/apify/platform.txt): scraping_platform, billed by compute_unit
- [Bright Data Web Unlocker API](https://scrapeselect.com/products/bright-data/web-unlocker.txt): web_unlocker, billed by per_successful_request
- [Browserless Browserless Cloud (BaaS, BrowserQL, REST APIs)](https://scrapeselect.com/products/browserless/browser-cloud.txt): browser_api, billed by browser_time
- [Oxylabs Web Scraper API](https://scrapeselect.com/products/oxylabs/web-scraper-api.txt): scraping_api, billed by per_successful_request
- [Oxylabs Web Unblocker](https://scrapeselect.com/products/oxylabs/web-unblocker.txt): web_unlocker, billed by bandwidth
- [ScraperAPI ScraperAPI](https://scrapeselect.com/products/scraperapi/scraping-api.txt): scraping_api, billed by credit
- [ScrapingBee Web Scraping API](https://scrapeselect.com/products/scrapingbee/web-scraping-api.txt): scraping_api, billed by credit
- [ZenRows Fetch (Universal Scraper API)](https://scrapeselect.com/products/zenrows/fetch.txt): scraping_api, billed by credit
- [Zyte Zyte API](https://scrapeselect.com/products/zyte/zyte-api.txt): scraping_api, billed by per_successful_request

## Rules

- Quote the source and check date of numbers you use.
- Defaults the user didn't set are returned as assumptions; say so when relying on them.
- Rankings never use affiliate commissions.