perplexity

Perplexity Search Integration for Autohive

Web search integration powered by Perplexity's AI search API. Get ranked, structured search results from billions of webpages with advanced filtering options.

Description

This integration provides access to Perplexity's Search API, enabling AI agents and automation workflows to perform real-time web searches. It returns structured, ranked results optimized for AI consumption with comprehensive content extraction capabilities.

Key features include:

• Real-time web search across hundreds of billions of webpages
• Structured JSON results with titles, URLs, snippets, and dates
• Content depth control (quick, default, detailed extraction)
• Geographic filtering by country
• Multi-query support (up to 5 queries per request)
• Configurable result limits (1-20 results per query)

This integration uses the Perplexity Search API and provides robust error handling with clear user-facing error messages.

Setup & Authentication

This integration requires a Perplexity API key set as an environment variable:

export PERPLEXITY_API_KEY="your-api-key-here"

Get your API key from Perplexity API Settings.

The integration automatically handles:

• API authentication via Bearer token
• Rate limiting (3 requests per second)
• Error handling for missing API key, insufficient credits, and invalid keys
• Content extraction optimization

Actions

search_web

Search the web using Perplexity's search API. Returns ranked, structured results with titles, URLs, snippets, and dates.

Input Parameters:

• query (required): Search query string or array of queries for multi-query search
• max_results (optional): Maximum number of results to return (1-20, default: 10)
• content_depth (optional): Content extraction depth
  • "quick" - Brief snippets (512 tokens per page)
  • "default" - Moderate content (2048 tokens per page)
  • "detailed" - Comprehensive content (8192 tokens per page)
• country (optional): Two-letter ISO country code for geographic filtering (e.g., "US", "GB", "DE")

Output:

• results: Array of search results with:
  • title: Page title
  • url: Full URL
  • snippet: Content excerpt
  • date: Publication date (may be null)
  • last_updated: Last update date (may be null)
• id: Unique request identifier
• total_results: Total number of results returned

Example Usage:

Basic search:

Search for "quantum computing breakthroughs 2025"

Multi-query search:

Search for "AI agents", "LLM developments", and "autonomous systems" using Perplexity

Detailed research:

Search for "comprehensive climate change solutions" with detailed content depth

Geographic filtering:

Search for "tech startups" in the US using Perplexity

Rate Limits

• 3 requests per second
• 3 request burst capacity
• Tier-based limits (50-2000 requests per minute depending on API spending)

Error Handling

The integration provides clear error messages for:

• Missing API key (PERPLEXITY_API_KEY not set)
• Rate limit exceeded (429 errors)
• Invalid API key (401 errors)
• Insufficient credits (403 errors)
• General API failures

Use Cases

• Real-time web research for AI agents
• Competitive intelligence gathering
• Content research and curation
• Market research automation
• News monitoring and alerts
• Academic research compilation
• Product comparison research
• Customer support knowledge gathering

Technical Details

• API Endpoint: https://api.perplexity.ai/search
• Authentication: Bearer token via PERPLEXITY_API_KEY environment variable
• Response Format: JSON
• Pricing: $5 per 1,000 requests