search1api-mcp

Search1API MCP Server is a versatile MCP server designed to enhance AI models with comprehensive search and content extraction capabilities. It empowers AI models to access real-time information through tools like web search, news search, and website crawling. Seamlessly integrated with MCP clients like Claude Desktop and Cursor, it offers functionalities such as web content extraction, sitemap retrieval, and even advanced reasoning powered by DeepSeek R1.

Developers benefit from its ability to solve complex problems by providing AI models with up-to-date information. The server supports various search services, including Google, Bing, and Reddit, and allows for refined searches using parameters like time range and site inclusion/exclusion. By simply configuring the server with a Search1API key, developers can unlock a wealth of knowledge for their AI models, enabling more informed and context-aware responses. It utilizes standard input/output for communication, ensuring easy integration into existing AI workflows.

Web Search Functionality

The search tool is the core of search1api-mcp, enabling AI models to access and process real-time information from the web. It allows models to formulate search queries in natural language and retrieve relevant results from various search engines like Google, Bing, DuckDuckGo, and specialized platforms like Reddit, GitHub, and YouTube. The tool offers parameters to refine searches, including the number of results, specific search services, and time ranges. This functionality empowers AI models to answer complex questions, conduct research, and stay updated on current events.

For example, an AI model tasked with writing a report on the latest advancements in AI could use the search tool with the query "recent AI breakthroughs" and a time range of "month" to gather up-to-date information. The crawl_results parameter can be used to extract the full content of the most relevant results, providing the model with in-depth information for its report. The ability to specify include_sites and exclude_sites allows for even more refined searches, focusing on authoritative sources and filtering out irrelevant content.

Web Page Content Extraction

The crawl tool allows AI models to extract the full content from specified URLs. This is crucial for scenarios where a summary or snippet from a search result is insufficient, and the model needs the complete context of a web page. By providing a URL, the tool retrieves the text, HTML, and other relevant data from the page, making it available for the AI model to analyze and utilize.

Consider an AI model designed to analyze customer reviews for a specific product. The model could use the search tool to find relevant review pages on different e-commerce websites. Then, using the crawl tool, it could extract the full content of each review page, enabling it to perform sentiment analysis, identify common issues, and provide valuable insights to the product development team. This feature significantly enhances the model's ability to understand and process information from diverse online sources.

Deep Reasoning with Web Search

The reasoning tool combines the power of a fast reasoning model (DeepSeek R1 by default, but configurable) with web search capabilities. This allows AI models to tackle complex problems that require both logical inference and access to external information. The tool takes a question or problem as input and uses the reasoning model to break it down into smaller steps, leveraging web search to gather relevant data for each step.

For instance, an AI model could use the reasoning tool to plan a trip to a new city. The model could start by searching for popular attractions, then use the crawl tool to extract detailed information about each attraction, including opening hours, ticket prices, and user reviews. Finally, the model could use the reasoning model to create an optimal itinerary based on the user's preferences and constraints. This integration of reasoning and web search enables AI models to solve complex, real-world problems more effectively.