Choose the right product
Start with the work you need to complete and the result your application needs. A browser session, page content, an AI answer, and a search result each call for a different starting point.
Match your task
| Your task | Start with | You provide | You work with |
|---|---|---|---|
| Navigate a site, fill forms, click elements, or run a session-based workflow | Agent Browser | Browser automation code and session configuration | A browser session controlled by your code |
| Collect one page, a list of pages, or linked pages across a website | Crawl | URLs and scrape or crawl options | Page content in selected formats, such as Markdown, HTML, and links |
| Collect answers and available citations from supported AI engines | AI Scraper | An engine selection, prompts, and supported parameters | Completed task results for that engine |
| Retrieve content from a URL with managed access handling | Web Unlocker | A target URL and request or rendering options | Content to parse or process in your application |
| Add IP routing to a collector or browser you already run | Proxies | Your traffic and proxy configuration | Requests routed through the selected proxy network |
| Get Google search results | Scraping API: Google Search API | A query and search parameters | Structured search results |
| Collect data from Amazon | Scraping API: Amazon Scraper | Inputs for the selected Amazon scraping action | Structured Amazon data |
| Collect data from TikTok | Scraping API: TikTok Scraper | Inputs for the selected TikTok scraping action | Structured TikTok data |
| Let an AI agent call multiple Scrapeless products | Scrapeless MCP | Tasks issued through an MCP client | Results from the selected product tools |
Decide between similar options
Agent Browser or Crawl?
Choose Agent Browser when your program needs to decide what to do next inside a browser: inspect an element, navigate after a click, or continue an authenticated workflow.
Choose Crawl when you can describe the collection job with URLs and options. For example, collecting documentation pages for a retrieval pipeline usually starts with Crawl; automating a multi-step portal workflow starts with Agent Browser.
Crawl or Web Unlocker?
Choose Crawl for a managed content collection workflow, especially when you need to process a URL list or discover linked pages across a site.
Choose Web Unlocker when your application already decides which URL to request and how to process the response. Its JavaScript rendering options also support request-level instructions and output formats. The choice is about the workflow you want to manage, rather than simply whether the page uses JavaScript.
Web Unlocker or Proxies?
Choose Web Unlocker to delegate supported access handling along with the content request.
Choose Proxies when you want to retain your existing browser or HTTP client and manage its rendering, parsing, and retry logic yourself. A proxy routes traffic; your application still determines how to collect and interpret the page.
AI Scraper or Scraping API?
Choose AI Scraper when your input is a prompt and you want an answer from a supported AI engine, such as ChatGPT or Perplexity.
Choose Scraping API when you want structured data from Google Search, Amazon, or TikTok. Start with Google Search API, Amazon Scraper, or TikTok Scraper for the source you need. For an AI answer experience, check the corresponding AI Scraper engine guide and its supported fields.
Amazon Scraper belongs to Scraping API. Amazon Alexa, listed under AI Scraper, is an AI engine option with a different purpose.
Engine and scraper support are specific to each guide. Check that your required source and fields are documented before building around them.
Choose your integration method
After selecting a product, choose an interface that fits your application:
| How you build | Starting point |
|---|---|
| You want to call endpoints directly from any language | API Reference |
| You use Node.js or TypeScript and want SDK methods | Node.js SDK |
| You use Python and want SDK methods | Python SDK — Coming soon |
| You use Go and want SDK methods | Go SDK — Coming soon |
| You want an AI agent to call supported Scrapeless tools | Scrapeless MCP |
| You prefer a command-line interface | Scrapeless CLI — Coming soon |
| You want to connect an existing workflow or agent tool | Integrations catalog |
Python and Go SDKs are in development. The CLI is also reserved as a Coming soon entry. You can call the REST APIs directly from either language while those SDKs are being prepared.
An SDK or MCP server provides an interface to supported capabilities. Verify its current methods or tools for your chosen product.
Try a small representative task
Begin with one URL, one prompt, or one search query. Check the returned fields and whether the result meets your needs before expanding to a larger job.
If you want to learn the basic authentication and request flow first, follow the platform Quickstart.