# AdsCrawl Official Documentation Index

> An English index of AdsCrawl's official API documentation, provided by the AdsCrawl website for developers and AI tools.

The documentation site at https://docs.adscrawl.net is the authoritative reference for API contracts, authentication, supported parameters, errors and current limits. This index summarizes the product boundaries and links to the relevant documentation; it does not replace those contracts.

## Start here

- [API introduction](https://docs.adscrawl.net/api-reference/introduction): API overview and authentication.
- [Quickstart](https://docs.adscrawl.net/quickstart): Make your first request.
- [API schemas](https://docs.adscrawl.net/api-reference/schemas): Shared request and response schemas.
- [Public OpenAPI specification](https://www.adscrawl.net/openapi.yaml): Machine-readable HTTP API contract.

The production API base URL is https://api.adscrawl.net. Send the full API key in the `x-api-key` header and keep it on your server. An account login session is not an API key. Remote CDP connection URLs carry their own scoped data token; do not replace it with the account API key.

## HTTP browser tasks

- [Browser tasks](https://docs.adscrawl.net/concepts/browser-tasks): Request-based browser rendering and extraction.
- [HTML, Markdown and article JSON](https://docs.adscrawl.net/api-reference/html): `POST /html` renders a page. `contentMode` selects rendered HTML or Readability-based article Markdown/JSON. Article JSON is not arbitrary schema extraction.
- [Structured extraction](https://docs.adscrawl.net/api-reference/spa-extract): `POST /spa-extract` inspects or extracts fields from the DOM and matching network responses. Use the documented field definitions or published templates.
- [Screenshots](https://docs.adscrawl.net/api-reference/screenshot): `POST /screenshot` returns PNG bytes. Capture the viewport, full page or first element matching a CSS selector; `selector` takes precedence over `fullPage`.

These APIs return task outputs. Their supported wait conditions, cookies, proxy settings and timeouts are endpoint-specific.

## Temporary Remote CDP

- [Remote CDP concepts](https://docs.adscrawl.net/concepts/remote-cdp): Temporary browser sessions for scripted interaction.
- [CDP session API](https://docs.adscrawl.net/api-reference/cdp-sessions): Create and release a remote Chromium session. Use the returned `cdpBaseUrl` directly with Playwright `connectOverCDP`, or follow the documented discovery flow for Puppeteer.
- [Playwright automation guide](https://docs.adscrawl.net/guides/playwright-automation): Connect an automation client and run browser actions.

Temporary CDP sessions have bounded lifetimes and idle limits. Release a session with `DELETE /cdp/sessions/{sessionId}` when the work is complete.

## Saved Cloud Browser environments

- [Cloud Browser concepts](https://docs.adscrawl.net/concepts/cloud-browsers): Saved profiles, runtime lifecycle and viewer access.
- [Cloud Browser session guide](https://docs.adscrawl.net/guides/cloud-browser-sessions): Create, start, reuse and stop cloud browser profiles.
- [Cloud Browser launch API](https://docs.adscrawl.net/api-reference/cloud-browsers-launch): Create a profile and start its runtime in one operation.

AdsCrawl Cloud Browser runs AdsPower fingerprint browsers with real native browser fingerprints. Each profile has an isolated environment with separate fingerprint settings and browser state. Cloud Browser follows AdsPower fingerprint browser version updates. These advantages describe Cloud Browser, not every HTTP task or temporary Remote CDP session.

Starting Cloud Browser runtime requires a paid plan. Launch requests require an explicit custom proxy; API-key requests that start an existing profile also require an explicit valid top-level proxy. Saved proxy settings do not replace that API-key start requirement. Supply proxy credentials in the designated fields, not inside the proxy server URL.

Stopping a runtime preserves its saved profile and snapshots. Closing the viewer does not stop runtime or billing: call the stop endpoint and confirm `runtime.status` is `stopped`. The live viewer requires the owner's login session. A saved profile does not guarantee a CDP endpoint; use a CDP URL only when the runtime returns one.

## Website resources

- [Product overview](https://www.adscrawl.net/): Choose an AdsCrawl workflow.
- [Cloud Browser API](https://www.adscrawl.net/cloud-browser-api/): AdsPower-based saved cloud environments.
- [Pricing](https://www.adscrawl.net/pricing/): Current plans and usage model; account-specific limits are shown in the dashboard.
- [Free API request generator](https://www.adscrawl.net/tools/api-request-builder/): Build HTML, Markdown or screenshot examples in cURL, JavaScript or Python. Generation runs locally; the tool does not execute requests or require an API key.
- [Frequently asked questions](https://www.adscrawl.net/faqs.md): Concise integration and product-selection answers.
- [Website index for LLMs](https://www.adscrawl.net/llms.txt): Product pages and related resources.