> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# TavilySearch Tool
> Guide for using the TavilySearch tool with PraisonAI agents.
The TavilySearch tool lets an agent run AI-optimised web search through the Tavily API.
```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
subgraph "Tavily Search"
User[User] --> Agent[Agent]
Agent --> Tool[TavilySearchResults]
Tool --> API[Tavily API]
API --> Result[Results]
end
classDef input fill:#8B0000,stroke:#7C90A0,color:#fff
classDef process fill:#189AB4,stroke:#7C90A0,color:#fff
classDef output fill:#10B981,stroke:#7C90A0,color:#fff
class User,Agent input
class Tool,API process
class Result output
```
## Overview
The TavilySearch tool is a tool that allows you to search the web using the TavilySearch.
```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
pip install "langchain-community>=0.2.11" tavily-python
export TAVILY_API_KEY="${TAVILY_API_KEY:?Set TAVILY_API_KEY in your shell}"
```
```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonaiagents import Agent, AgentTeam
from langchain_community.tools import TavilySearchResults
def search_tool(query: str):
tool = TavilySearchResults(
max_results=5,
search_depth="advanced",
include_answer=True,
include_raw_content=True,
include_images=True
)
return tool.run(query)
data_agent = Agent(instructions="I am looking for the top google searches on AI tools of 2025", tools=[search_tool])
editor_agent = Agent(instructions="Analyze the data and rank the tools based on their popularity")
agents = AgentTeam(agents=[data_agent, editor_agent])
agents.start()
```
## How It Works
```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
sequenceDiagram
participant User
participant Agent
participant Tavily as TavilySearchResults
participant API as Tavily API
User->>Agent: "Search for X"
Agent->>Tavily: run(query)
Tavily->>API: HTTPS request
API-->>Tavily: results + answer
Tavily-->>Agent: parsed result
Agent-->>User: final answer
```
## Getting Started
1. Install dependencies (see **Overview** above)
2. Set required API keys in your environment
3. Run the agent example in **Overview**
Use the same tool with an agent — see the **Overview** example, or pass env vars from the sections above.
## Best Practices
Set `TAVILY_API_KEY` in your shell or `.env`. The tool reads it automatically — never hard-code the key.
`TavilySearchResults(max_results=5)` controls how many hits return. Lower it so the agent processes fewer tokens.
Set `include_answer=True` when the agent needs a synthesised answer rather than raw links — it saves a follow-up LLM step.
## Related Tools
Tavily tool (praisonai-tools)
Exa LangChain integration
Privacy-focused search