# Introduction

> Search1API gives AI applications one API for discovering, reading, and structuring information from the live web.

Canonical URL: https://s1.dev/docs

Search1API gives AI applications live web access through one API. It combines web and news search, page crawling, webpage screenshots, site discovery, trending feeds, and schema-driven extraction behind the same base URL and API key.

Use it when an agent, RAG pipeline, or backend needs current information without maintaining search-engine adapters or scraping infrastructure.

## A simple mental model

- [Discover](https://s1.dev/docs/basic/search): Find relevant pages with web search, describe what you need in plain language with agentic search (Ask), get recent coverage with news search, or follow current developer topics with trending feeds.

- [Read](https://s1.dev/docs/basic/crawl): Turn one page into clean text, fetch the top search results with crawl_results, or crawl a site asynchronously.

- [Capture](https://s1.dev/docs/basic/screenshot): Render a full page, viewport, or page element as PNG, JPEG, or WebP.

- [Structure](https://s1.dev/docs/advanced/extract): Extract fields from a page into a JSON shape you define.

Most agent workflows begin with `/search`. Add `crawl_results` when the model needs the source text rather than links and snippets. If you already have a URL, start with `/crawl` instead. If you know what you want but not which engine to ask, use [`/ask`](https://s1.dev/docs/basic/ask) and let Search1API choose the engines and time window.

For the complete decision guide, see [Choosing an endpoint](https://s1.dev/docs/guides/choosing-an-endpoint).

## Choose your starting point

- [Make your first API call](https://s1.dev/docs/get-started/quickstart): Create a key, run one request, and understand the response.

- [Add an official SDK](https://s1.dev/docs/integrations/sdks): Install the TypeScript or Python client for application code.

- [Connect an AI tool](https://s1.dev/docs/integrations/mcp): Use Search1API through MCP, the CLI, or an agent skill.

- [Choose an endpoint](https://s1.dev/docs/guides/choosing-an-endpoint): Match search, crawl, screenshot, extract, sitemap, and deepcrawl to your task.

## When you are ready for production

- [Authentication](https://s1.dev/docs/essentials/authentication): Store, separate, and revoke API keys safely.

- [Credits and limits](https://s1.dev/docs/essentials/credits-and-limits): Understand request costs, balances, and rate limits.

- [Error handling](https://s1.dev/docs/essentials/error-handling): Handle validation, authentication, payment, and upstream failures.

- [Changelog](https://s1.dev/docs/changelog): Compatibility promise and dated changes to the API, SDKs, CLI, and MCP server.