# Perplexity AI MCP for AI Agents AI Agent Connect

> Perplexity AI MCP gives your agent real-time web search capabilities. It lets your AI client browse the live internet to find current news, academic papers, and technical data, providing grounded answers with direct citations for every claim it makes.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_lW0xnD7wkhoOhFgYCfH30bKS54sVQUOFu0ibRfR1/ai-agent-connect
- **Tags:** web-search, research-assistant, citations, natural-language, information-retrieval

## Description

You can connect Perplexity AI to your agent to give it live eyes on the internet. Instead of your agent guessing based on old training data, it pulls current facts, news, and research papers. It handles the heavy lifting of browsing, synthesizing the info, and providing the sources so you don't have to double-check every claim. This is a major jump in reliability for anyone who needs to know what happened ten minutes ago, not two years ago. You'll find this in the Vinkius catalog alongside other high-performance tools. It turns your agent into a research partner that actually backs up its work with links, making it much easier to trust the output for professional work.

## Tools

### chat_completion
Get a quick, grounded answer to a general knowledge question with citations. Use this for fast fact-checking.

### chat_with_citations
Receive answers where every single claim is linked back to its original source. This is the core of reliable research.

### chat_with_images
Get relevant images along with your text answers for visual topics. It's helpful for product searches or visual identification.

### chat_with_domain_filter
Limit your agent to searching only specific websites like research journals or news outlets. This ensures high-quality sources.

### chat_with_history
Keep the context of a long conversation so you can ask follow-up questions easily. It makes multi-turn research feel natural.

### chat_with_recency_filter
Filter search results by timeframes like hours or days to get only the latest news. This keeps your agent from pulling outdated information.

### chat_with_related_questions
Get a list of follow-up questions to help you explore a topic more deeply. It's great for starting new research branches.

### deep_research
Generate a long, detailed report on complex subjects with exhaustive web searching. It takes longer but provides much more depth.

### follow_up
Continue a multi-turn research session by building on previous answers. It maintains the thread of your conversation.

### reasoning
Solve math problems or analyze code with step-by-step logic and citations. It handles complex reasoning tasks well.

### search_query
Get the full package of citations, images, and related questions in one go. It's the most complete way to search the web.

### structured_query
Get your results in a specific JSON format that matches your own schema. This is perfect for programmatic data extraction.

### system_prompt_query
Give your agent a specific persona or set of rules to follow for the session. Use it to set a specific expert tone.

### list_models
See which specific models are available for your queries before you start. This lets you choose the right tool for the job.

## Prompt Examples

**Prompt:** 
```
What are the current interest rates for small business loans in the US today?
```

**Response:** 
```
### Current Small Business Loan Interest Rates (US)

| Lender Type | Average Rate | Key Features |
| :--- | :--- | :--- |
| **Online Lenders** | 7.9% - 14.5% | Fast approval, flexible terms |
| **Community Banks** | 6.5% - 11.0% | Relationship-based, lower rates |
| **SBA Loans** | 6.0% - 10.5% | Government-backed, longer terms

**Key Trends:**
* Rates have stabilized slightly this month but remain higher than 2021 levels.
* Online lenders are offering 'fast-track' options for loans under $50k.

📚 **Sources:** Federal Reserve, SBA.gov, Bankrate.com
```

**Prompt:** 
```
Give me a summary of the latest AI regulations passed in the EU this month.
```

**Response:** 
```
### EU AI Regulation Update - Current Month

**Key Developments:**
* **Implementation Guidelines:** New clarity on how the AI Act applies to 'high-risk' systems in healthcare.
* **Office Establishment:** The EU AI Office has begun onboarding staff for enforcement.
* **Compliance Deadlines:** Member states have finalized their national transition plans.

**Impact:** Companies must now begin auditing their existing models for bias and transparency.

📚 **Sources:** Europa.eu, Reuters, TechCrunch
```

**Prompt:** 
```
Find the latest documentation for the Stripe API's new checkout features.
```

**Response:** 
```
### Stripe API: New Checkout Features

Stripe recently updated their Checkout API with several key additions:
* **Multiple Payment Methods:** Support for local payment methods in 15+ new countries.
* **Dynamic Shipping:** Real-time shipping rate calculation at the start of checkout.
* **Tax Automation:** Integrated tax calculation for cross-border sales.

**Action Items:**
1. Update your `checkout.sessions.create` calls.
2. Review the new `shipping_options` object.

📚 **Sources:** Stripe Documentation, Developer Blog
```

## Capabilities

### Get cited answers
Your agent provides links to the original sources for every fact it shares.

### Restrict search to specific sites
Limit your agent to searching only trusted domains like government or academic sites.

### Filter by recency
Tell your agent to only look for information from the last hour, day, or month.

### Generate detailed reports
Ask for a deep dive into a topic and get a long form report with exhaustive searching.

### Extract structured JSON
Get web data returned in a specific format that your other apps can read easily.

## Use Cases

### Market Trend Analysis
An analyst asks for a summary of competitor pricing from the last 48 hours. The agent uses chat_with_recency_filter to find the latest data.

### Academic Literature Review
A student needs to find recent papers on carbon capture. The agent uses chat_with_domain_filter to search only peer-reviewed journals.

### Technical Troubleshooting
A developer is stuck on a new API error. The agent uses chat_completion to find the latest community fixes and documentation updates.

### Fact-Checking for News
A writer needs to verify a quote from a government official. The agent uses chat_with_citations to provide the exact source link.

## Benefits

- Stop guessing what your agent knows. Perplexity AI MCP ensures your agent uses live web data instead of outdated training info.
- Verify every claim with chat_with_citations. You get direct links to sources so you can double-check the facts yourself.
- Control your sources with chat_with_domain_filter. You can force your agent to only look at trusted sites like .gov or .edu.
- Get deep dives with deep_research. Move past surface-level answers to get long form reports on complex, multi-faceted topics.
- Keep your data organized with structured_query. Get web results in a clean JSON format that you can pipe directly into other apps.
- Stay current with chat_with_recency_filter. Tell your agent to only look for news from the last hour or day to stay up to date.

## How It Works

The bottom line is your agent gets a real-time research engine that cites its sources.

1. Subscribe to the Connector and grab your Perplexity API key from your account settings.
2. Add the connection to your AI client like Claude or Cursor.
3. Ask your agent to research a topic and it will return cited answers with links.

## Frequently Asked Questions

**Can Perplexity AI MCP find news from the last hour?**
Yes. It can filter results by recency, allowing your agent to pull information from the last hour, day, or week. This is perfect for staying on top of breaking news.

**How do I get sources from Perplexity AI MCP?**
Every answer provided by your agent will include a citations array. This means you'll see direct links to the websites and papers the agent used to find the information.

**Can Perplexity AI MCP search only specific websites?**
Yes. You can tell your agent to only look at specific domains. This is great for ensuring your agent only uses trusted sources like official government sites or academic journals.

**Will Perplexity AI MCP help with academic research?**
It's built for it. It can perform deep research on complex topics and provide long form reports with thorough citations, making it much easier to compile bibliographies and literature reviews.

**Can I get the data from Perplexity AI MCP in a format I can use in my code?**
Yes. You can ask your agent to provide a structured JSON response. This allows you to get web data in a specific format that you can easily parse and use in your own applications.

**Is Perplexity AI MCP good for complex math or logic?**
Yes. It includes a specific reasoning capability that helps your agent solve multi-step math problems and perform logical deductions while still citing its sources.

**How do I get a Perplexity API key?**
Log in to your [**Perplexity AI account**](https://www.perplexity.ai/), go to **Settings > API**, and generate a new API key. Copy the key (it starts with `pplx-`) immediately. Paste it into the API key field below. This key authenticates all requests to https://api.perplexity.ai.

**What's the difference between Sonar, Sonar Pro, Deep Research, and Reasoning Pro models?**
**Sonar** is the fastest model for quick factual answers and basic synthesis. **Sonar Pro** handles complex queries better with more thorough analysis and follow-up support. **Sonar Deep Research** performs exhaustive web searches and generates comprehensive reports with thorough citations — best for research papers and deep investigations. **Sonar Reasoning Pro** excels at logical reasoning, multi-step analysis, mathematical problems, and chain-of-thought tasks.

**Can I restrict search results to specific domains or time periods?**
Yes! Use `chat_with_domain_filter` to restrict search to specific domains (e.g., arxiv.org, nih.gov, github.com). Use `chat_with_recency_filter` to get results only from the last hour, day, week, month, or year. You can also combine both for domain-specific recent information. Citations are automatically included to verify sources.

**How does Perplexity AI differ from regular search engines?**
Unlike regular search engines that return a list of links, Perplexity AI reads the web in real-time, synthesizes information from multiple sources, and provides a direct, concise answer with citations. It's like having a research assistant that reads dozens of pages and summarizes the key findings with source links. You get answers, not just links.