← All templatestemplates / web-research-agent-with-real-time-search-and-content-extraction

Cartesiaupdated May 28, 2025 · other · voice-search

Web Research Agent with Real-Time Search and Content Extraction

A voice agent that searches the web and extracts page content using the Tavily API, then synthesizes results into conversational responses. Uses Tavily's fast search depth for low-latency voice interactions and deep content extraction for specific pages.

Try the demo View source Fork template

The numbers

latency—

cost / min—

frameworkCartesia

The stack

telephonyWeb Only

speech-to-textCartesia STT

llmGPT-4o

text-to-speechCartesia Sonic-3

System prompt

No prompt published.

Config

config.json

{
  "temperature": 0.7,
  "tools": [
    {
      "name": "web_search",
      "description": "Calls Tavily Search API and returns formatted results to the LLM. Uses fast search depth for low-latency voice interactions with max 5 results."
    },
    {
      "name": "web_extract",
      "description": "Extracts full content of a webpage by URL using Tavily Extract API. Content is truncated to 3000 characters for LLM context management."
    }
  ],
  "max_tokens": 600,
  "tavily_max_results": 5,
  "tavily_search_depth": "fast",
  "extract_truncate_chars": 3000
}

Tools

web_searchCalls Tavily Search API and returns formatted results to the LLM. Uses fast search depth for low-latency voice interactions with max 5 results.
web_extractExtracts full content of a webpage by URL using Tavily Extract API. Content is truncated to 3000 characters for LLM context management.

Voice AI recipes, picks, and analysis.

Get the useful new templates plus the occasional teardown of what’s working in production voice AI.