A developer-first alternative to Kadoa for web scraping
If you need a controllable scraping API that returns clean HTML, JSON, Markdown or plain text, this page explains the practical choices: direct API calls, Proxy Mode for proxy-expecting tools, JSON capture from dynamic apps, element screenshots, and a dedicated search endpoint.
Use the checklist below to match the workflow to your stack before committing to a plan.
Start with a direct API call
The REST workflow is simple: send a URL and receive rendered HTML or structured JSON, with options for rendering, extraction and screenshots. Headless browsers and rotating proxies are handled for you, so your team can focus on parsing and using the returned data rather than maintaining browser infrastructure.
Use Proxy Mode when your tool expects a proxy
Some tools cannot call a REST API but can use a proxy endpoint. Proxy Mode acts as a forward-proxy front-end for those tools, while keeping rendering and extraction controls available. Decide based on the client you already have: API client for new code, proxy endpoint for legacy or third-party software.
Match the output to where the data lives
XHR/Ajax interception returns the JSON a dynamic app loads, so you can extract data where it actually lives instead of parsing the initial HTML. For visual checks, screenshot_selector captures full pages or specific elements; for indexing and AI pipelines, Markdown or plain-text export replaces raw HTML. A worked example: if a product page renders prices through a JSON call after load, request the JSON response rather than parsing the rendered HTML; if you need to verify a badge visually, request an element screenshot; if you are feeding a retrieval pipeline, request Markdown.
Keep search data on its own endpoint
The source product describes a dedicated Google Search endpoint that returns realtime SERP data for SEO monitoring and keyword research, separate from general page scraping. Keep search workflows on that endpoint and general extraction on the main API so each job uses the format it needs.
Control, debug and track each request
Per-request controls let you forward headers, pass cookies, keep sessions and tune rendering without adding infrastructure or maintaining browsers. Response metadata such as credit cost and resolved URL makes monitoring spending and debugging failures at scale straightforward. Clear parameter documentation and quick examples shorten the path from first request to a stable workflow.
Decision checklist before you commit
Before you choose a plan, run this check: 1) Pick the output you need — HTML, JSON, Markdown or plain text. 2) Decide API or Proxy Mode based on your client. 3) Test a dynamic page with XHR interception if the data is not in HTML. 4) Estimate credits using the per-request costs for rendering and Premium proxies. 5) Confirm current plan allowances and terms with the owner before budgeting, since prices and offers can change.
| Workflow | Best fit | Output to request |
|---|---|---|
| Direct REST API | New code that can send a URL and parse a response | HTML or structured JSON |
| Proxy Mode | Existing tool that expects a proxy endpoint | Rendered page via proxy |
| XHR/Ajax interception | Dynamic app where data loads as JSON | JSON response data |
| Element screenshot | Visual verification of a specific element | Screenshot image |
| Markdown / plain text | Indexing or AI workflows needing clean content | Markdown or plain text |
| Dedicated search endpoint | SEO monitoring and keyword research | Realtime SERP data |
Use this as a selection checklist; confirm current plan terms separately.
Frequently asked questions
When should I use the API instead of Proxy Mode?
Choose the API route when your code can send a URL and parse HTML, JSON, Markdown or plain text. Choose Proxy Mode when the tool you already use expects a proxy endpoint rather than an API client; rendering and extraction options remain available in that workflow.
When is JSON from XHR/Ajax the right output?
Use XHR/Ajax interception when the useful data is loaded as JSON by a dynamic application and is not present in the initial HTML. Use rendered HTML or extraction rules when the content you need is in the page itself.
How should I estimate credit use before choosing a plan?
Plan by request cost, not only by request count. Credits are described as valid for one month without rollover, requests start at 1 credit, JavaScript rendering costs 5 credits per successful request, and Premium proxies cost 10 or 25 credits depending on rendering. Confirm current plan allowances before budgeting.