# Perplexity AI MCP for AI Agents AI Agent Connect

> Perplexity AI MCP connects your Perplexity account to your AI agent to provide web-grounded answers with real-time citations. It allows your agent to search the live web, filter by domain, and use specialized models like Sonar Pro for deep research without you having to switch between different tabs or tools.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_UbSKRkwZQEYfv4pzyvcNwmqlKwk5UEhIITyylXnJ/ai-agent-connect
- **Tags:** web-search, ai-chat, citations, real-time-data, natural-language-processing, search-api

## Description

Perplexity AI connects your Perplexity account to any AI agent so you can get web-grounded answers with citations. It turns your AI client into a research powerhouse that actually knows what happened today. Instead of copying and pasting URLs into a separate browser tab or manually searching for news, you can let your agent do the heavy lifting. It pulls in live web results, provides direct links to sources, and lets you toggle between different models depending on whether you need a quick fact or a deep, multi-step logical breakdown. If you're already using the Vinkius catalog to manage your tech stack, adding this is the easiest way to give your agent a grounded brain. It handles the messy work of navigating search results so you can stay focused on the actual analysis. You get answers that are verifiable, cited, and specific to the domains you care about. No more switching between search engines and chat tools; your AI acts as a web-grounded research assistant. You can filter your results by specific domains to cut out the noise, ensuring your agent only looks at the sources you trust. Whether you are looking for a specific price point, a recent scientific breakthrough, or a deep dive into a niche industry, this connection provides the real-time data needed to make informed decisions without the risk of hallucinations.

## Tools

### chat
Send a message to a Perplexity model and receive a response with web citations. This helps you get grounded answers.

### chat_pro
Send a message to the Sonar Pro model for enhanced, high-quality responses with citations. Use this for better accuracy.

### chat_with_reasoning
Send a message to the Sonar Reasoning model for step-by-step logical breakdowns. This is great for complex problems.

### chat_with_reasoning_pro
Send a message to the Sonar Reasoning Pro model for deep reasoning with citations. This handles the most difficult tasks.

### get_usage
Retrieve your current API usage statistics to monitor your consumption limits. Use this to stay within your budget.

### list_models
List all available Perplexity models and their specific capabilities. Use this to see what models you can access.

### search
Perform a web search to get snippets, citations, and source URLs for a query. This provides real-time data.

### sonar
Send a message to the core Sonar model for web-grounded responses with citations. This is the main search model.

## Prompt Examples

**Prompt:** 
```
What are the top 3 news stories in renewable energy from this morning?
```

**Response:** 
```
### Top Renewable Energy News

* **Breakthrough in Solid-State Batteries**
  A new laboratory report shows a 20% increase in energy density. 
  [Source: ScienceDaily]

* **Grid Expansion Project Approved**
  The regional council approved a $2B expansion for wind farms. 
  [Source: Reuters]

* **Solar Efficiency Record**
  New perovskite cells hit a new efficiency peak. 
  [Source: Nature]

| Topic | Key Update | Source |
| :--- | :--- | :--- |
| Batteries | +20% Density | ScienceDaily |
| Grid | $2B Approval | Reuters |
| Solar | New Record | Nature |
```

**Prompt:** 
```
Find the current price of Ethereum and tell me the percentage change in the last 24 hours.
```

**Response:** 
```
### Ethereum (ETH) Market Update

**Current Price:** $2,450.12 USD
**24h Change:** +3.4%

**Summary:**
Ethereum is seeing a steady climb today, driven by increased activity in decentralized finance protocols. 

**Sources:**
* CoinMarketCap
* CoinGecko
```

**Prompt:** 
```
Explain the latest updates to the React documentation for server components.
```

**Response:** 
```
### React Server Components: Latest Updates

Recent updates focus on improving the **Request Waterfall** and **Data Fetching** patterns. Key changes include:

* **Improved Suspense Integration:** Better handling of nested components.
* **Request Memoization:** Automatic caching of fetch requests to prevent redundant calls.
* **Action Support:** Simplified way to handle mutations in server components.

Detailed documentation can be found on the official React website.
```

## Capabilities

### Get live web search results
Fetch real-time data with snippets and source URLs.

### Chat with cited sources
Receive answers from the Sonar models that include direct links to the web.

### Filter search by domain
Limit your agent's research to specific websites or regions.

### Access deep reasoning
Use the Sonar Reasoning models to solve complex, multi-step problems.

### Check API usage
Monitor your consumption limits to keep your workflow running smoothly.

### List available models
See every Perplexity model and its specific capabilities in one list.

## Use Cases

### Fact-checking a news story
Ask the agent to find the latest price of a commodity and provide sources using the search tool.

### Literature review
Use sonar to find recent papers on a specific niche topic with citations for a research project.

### Competitive analysis
Filter search results to a specific competitor's domain to see their latest updates and news.

### Debugging new tech
Ask for the latest documentation on a library released last week to ensure your code is up to date.

## Benefits

- Stop hallucinations: The sonar tools ensure your agent stays grounded in facts by pulling from the live web.
- Verify every source: Use the citations in chat_pro to jump directly to the original article or data point.
- Focus your research: Use search with domain filters to ignore the noise and only see results from trusted sites.
- Solve complex logic: Use chat_with_reasoning when you need the agent to think through a problem step-by-step.
- Save time on tabs: No more jumping between your AI client and a search engine; your agent handles both.
- Stay on budget: Use get_usage to keep an eye on your API spend while you work.

## How It Works

The bottom line is you get a web-connected research agent that cites every fact it finds.

1. Subscribe to the Connector and grab your Perplexity API key.
2. Plug that key into your AI client settings.
3. Ask your agent to search the web or reason through a complex topic.

## Frequently Asked Questions

**Does Perplexity AI MCP give my agent access to live news?**
Yes, it connects your agent to Perplexity's live search engine. This means your agent can find information about events happening right now, rather than relying on its training data.

**Can I limit the search to specific websites?**
Yes, you can use domain filters. This allows you to tell your agent to only look for information on specific websites you trust, which is great for professional research.

**Will my agent provide sources for its answers?**
Yes, the Sonar models are designed to provide web-grounded responses. Every fact it pulls from the web will include a citation and a source URL for you to verify.

**How do I know if I'm hitting my API limits?**
You can use the get_usage tool to check your current API consumption. This helps you monitor your spend and stay within your Perplexity account limits.

**Is this better than just searching the web manually?**
It's much faster for research. Instead of you browsing multiple sites, your agent does the searching, synthesizes the information, and gives you a structured summary with links.

**Can I use this for deep logical reasoning?**
Yes, the reasoning models are specifically built for this. They can think through complex, multi-step problems and provide a detailed chain of logic alongside their web-grounded findings.

**How do I get a Perplexity API key?**
Log in to the [**Perplexity AI Settings**](https://www.perplexity.ai/settings/api), go to **API Key Management** and click **Create API Key**. Copy the key immediately — it starts with `pplx-` and won't be shown again.

**What models are available?**
Use the `list_models` tool to see all available Perplexity models. Key models include sonar (web-grounded chat), sonar-pro (enhanced reasoning), sonar-reasoning and sonar-reasoning-pro. Each model provides real-time web search with citations.

**Can I filter search results to specific domains?**
Yes! Use the `chat` or `search` tools with the search_domains parameter to limit results to specific websites (e.g. ["wikipedia.org","github.com"]). This is useful for research focused on trusted sources.

**Do responses include citations?**
Yes! All Sonar models return responses with inline citations and a list of source URLs. The citations are embedded in the response text as numbered references [1], [2], etc., and the full source URLs are provided in the response metadata.