Exa AI

Comprehensive guide to the Exa AI node in StackAI: Learn how to use Exa AI for internet-scale search, including action details, input/output parameters, and practical examples.

What is Exa AI?

Exa AI is a powerful node in StackAI that enables users to perform advanced, internet-scale searches using both embeddings-based and traditional search methods. It allows you to query a wide variety of sources, retrieve relevant results, and integrate real-time web data into your AI workflows.


How to use it?

To use the Exa AI node in StackAI, simply add the node to your workflow and select the desired action. Configure the required input parameters to define your search query and any additional options. The node will return structured search results that can be used in downstream nodes for further processing, analysis, or display.


Example of Usage

Suppose you want to perform a web search for the latest AI research papers. You would select the "Web Search" action, provide your query (e.g., "latest AI research papers"), and configure any optional parameters such as the number of results. The node will return a list of relevant web pages, including titles, URLs, and summaries.


Available Actions in Exa AI

Below are the most commonly used actions available in the Exa AI node:


Description: Performs a real-time web search using advanced algorithms to retrieve the most relevant results from the internet.

Inputs:

  • query (string, required): The search term or question you want to look up.

    • Example: "latest AI research papers"

  • num_results (integer, optional): Number of search results to return.

    • Example: 5

Configurations: No additional configurations are required for this action.

Outputs:

  • results (array, required): List of search results, each containing:

    • title (string): Title of the web page.

    • url (string): Direct link to the web page.

    • snippet (string): Short summary or excerpt from the page.

Example:

{
  "results": [
    {
      "title": "Recent Advances in AI Research",
      "url": "https://example.com/ai-research",
      "snippet": "This article discusses the latest breakthroughs in artificial intelligence..."
    }
  ]
}

2. Deep Research

Description: Conducts an in-depth search and analysis on a given topic, aggregating information from multiple sources for comprehensive insights.

Inputs:

  • query (string, required): The topic or question for deep research.

    • Example: "impact of AI on healthcare"

  • num_results (integer, optional): Number of sources to aggregate.

    • Example: 3

Configurations: No additional configurations are required.

Outputs:

  • summary (string, required): A synthesized summary of findings.

  • sources (array, required): List of source URLs and brief descriptions.

Example:

{
  "summary": "AI is transforming healthcare by improving diagnostics, patient care, and operational efficiency...",
  "sources": [
    {
      "url": "https://example.com/ai-healthcare",
      "description": "Overview of AI applications in healthcare."
    }
  ]
}

3. Find Similar

Description: Finds web pages or documents similar to a provided URL or text snippet.

Inputs:

  • url (string, required): The URL of the reference page.

  • num_results (integer, optional): Number of similar results to return.

    • Example: 5

Configurations: No additional configurations are required.

Outputs:

  • similar_results (array, required): List of similar web pages with titles, URLs, and similarity scores.

Example:

{
  "similar_results": [
    {
      "title": "Understanding AI",
      "url": "https://example.com/understanding-ai",
      "score": 0.92
    }
  ]
}

4. Get Contents

Description: Retrieves the full content of a web page or document from a given URL.

Inputs:

Configurations: No additional configurations are required.

Outputs:

  • content (string, required): The extracted text content of the page.

Example:

{
  "content": "Artificial intelligence (AI) is a rapidly evolving field..."
}

5. Answer

Description: Provides a direct answer to a question by searching and synthesizing information from the web.

Inputs:

  • question (string, required): The question you want answered.

    • Example: "What is generative AI?"

Configurations: No additional configurations are required.

Outputs:

  • answer (string, required): The synthesized answer.

  • sources (array, required): List of source URLs used to generate the answer.

Example:

{
  "answer": "Generative AI refers to artificial intelligence systems that can create new content...",
  "sources": [
    "https://example.com/generative-ai"
  ]
}

Best Practices for Using Exa AI in StackAI

  • Always provide clear and specific queries for the best results.

  • Use the "num_results" parameter to control the amount of data returned.

  • Integrate Exa AI outputs with downstream nodes for advanced processing, such as summarization or visualization.

  • Review the sources and content for accuracy, especially when using results in critical applications.


Summary

The Exa AI node in StackAI is a versatile tool for integrating real-time, internet-scale search and research capabilities into your workflows. By leveraging its powerful actions, you can access, analyze, and utilize web data efficiently and effectively.

Last updated

Was this helpful?