# Perplexity AI MCP

> Perplexity AI MCP connects your agent to real-time web search capabilities. Get answers that are grounded in current information, complete with source citations and links. Stop relying on generalized knowledge; use this for deep research, fact-checking, and complex analysis directly from any compatible client.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Tags:** web-search, research-assistant, citations, natural-language, information-retrieval

## Description

This MCP turns your agent into a dedicated, academic research assistant. You stop manually switching between search engines, reading through dozens of results, and then trying to figure out which sources are trustworthy. Instead, you ask a question—whether it's about the latest quantum computing breakthroughs or the historical market share of electric vehicles—and get an answer synthesized from multiple real-time web sources. The AI doesn't just guess; it builds comprehensive answers while linking every claim back to its original source citation. This makes it indispensable for anyone who needs accuracy, whether you're writing a technical report or building a knowledge base in your application. Connecting this MCP through Vinkius gives your agent access to best-in-class research synthesis, making guesswork obsolete.

## Tools

### chat_completion
Ask a general question to get an answer with real-time web search and source citations for quick lookups.

### chat_with_citations
Get answers where every fact or claim is linked directly to its original published source URL, perfect for academic work.

### chat_with_domain_filter
Restrict the search results to specific domains you provide (like government sites) to ensure high-trust sources only.

### chat_with_history
Ask follow-up questions by providing a JSON array of previous messages so the model understands context.

### chat_with_images
Request visual results alongside text answers, which include URLs to relevant images found during the search.

### chat_with_recency_filter
Filter search results by time period (hour, day, week) so you only get information on recent events or news.

### chat_with_related_questions
Receive a list of suggested follow-up questions to help guide your research and discover related topics.

### deep_research
Run extensive, in-depth investigations that generate detailed reports with thorough citations on complex subjects.

### follow_up
Continue a research conversation by supplying the history of messages along with your next question.

### list_models
Check which specific search models are available before running your query to ensure you pick the right tool for the job.

### reasoning
Use this model for complex tasks like mathematical proofs, code reviews, or logical deductions requiring step-by-step analysis.

### search_query
Perform the most comprehensive search possible, combining citations, images, and related follow-up questions in one result.

### structured_query
Force the AI to return its answer as a JSON object that matches a specific data schema you define for programmatic use.

### system_prompt_query
Define a custom role or set of rules (e.g., 'Act like a financial auditor...') so the model adopts a specific persona and tone.

## Prompt Examples

**Prompt:** 
```
What are the latest developments in quantum computing as of this week?
```

**Response:** 
```
Based on recent developments this week, there are three major breakthroughs in quantum computing:

1. **IBM's 1000+ qubit processor** achieved new error correction milestones, reducing error rates by 50% compared to previous generations.
2. **Google Quantum AI** demonstrated quantum advantage in optimization problems for logistics, showing 100x speedup over classical algorithms.
3. **A team at MIT** developed a new topological qubit design that maintains coherence at higher temperatures.

📚 **Sources:** Nature.com, MIT News, IBM Research Blog

Would you like me to research any of these breakthroughs in more detail?
```

**Prompt:** 
```
Do deep research on the competitive landscape of electric vehicle manufacturers in Southeast Asia, including market share, pricing strategies, and government incentives.
```

**Response:** 
```
I've conducted comprehensive research on the EV competitive landscape in Southeast Asia. Here's the analysis:

## Market Leaders
1. **BYD** — 35% market share, aggressive pricing ($15K-30K), strong government partnerships in Thailand and Indonesia
2. **Tesla** — 18% market share, premium positioning ($40K+), expanding Supercharger network
3. **Neta (Hozon)** — 12% market share, budget-friendly ($10K-20K), popular in Vietnam

## Government Incentives
- **Thailand:** Up to ฿150,000 ($4,200) subsidy per EV, 0% import duty until 2025
- **Indonesia:** 0% luxury tax for EVs, $3,500 purchase incentive
- **Vietnam:** 0% registration fee until 2027, reduced road tax

## Key Strategies
- Local manufacturing to avoid import tariffs
- Battery swapping infrastructure partnerships
- Fleet sales to ride-hailing platforms (Grab, Gojek)

📚 **15 sources cited** including Reuters, BloombergNEF, and regional government reports.

Would you like me to dive deeper into any specific country or manufacturer?
```

**Prompt:** 
```
Search for news about AI regulation in the European Union from the last month, restricted to europa.eu and reuters.com domains.
```

**Response:** 
```
I found 5 recent articles about AI regulation in the EU from the last month, restricted to europa.eu and reuters.com:

1. **EU Commission publishes implementation guidelines for the AI Act** — europa.eu, 2 weeks ago
2. **European countries scramble to meet AI Act compliance deadlines** — reuters.com, 1 week ago
3. **EU establishes new AI Office to enforce regulations** — europa.eu, 3 weeks ago
4. **Industry leaders respond to EU AI Act risk classification system** — reuters.com, 4 days ago
5. **EU member states finalize national AI regulatory frameworks** — europa.eu, 5 days ago

📚 **Citations:** All 5 sources from europa.eu and reuters.com with direct links.

Would you like me to summarize any of these articles in detail?
```

## Capabilities

### Retrieve verifiable facts with sources
The system provides answers and links every claim back to its original source URL.

### Perform deep, exhaustive literature reviews
Run extended research queries that generate detailed reports covering complex topics and multiple angles.

### Focus searches on specific industries or academic domains
Restrict the search results to a list of trusted websites, like government sites or journals, eliminating irrelevant noise.

### Maintain context across multi-step research sessions
The agent remembers previous questions and answers, allowing you to refine complex queries naturally over time.

### Structure data into predictable formats
Force the AI to return results as JSON, making it simple for your code to ingest structured information immediately.

## Use Cases

### Comparing global EV market incentives
A market analyst needs to compare tax subsidies across three countries. They use `chat_with_domain_filter` to restrict searches only to national government sites, ensuring the pricing data is current and official.

### Verifying a historical claim for a paper
A student writes about quantum computing and needs proof. They run `chat_with_citations`, which guarantees that every technical detail they include has an immediate, verifiable source link attached to the answer.

### Extracting contact data from a webpage
A developer finds a complex article and needs only names and phone numbers. They use `structured_query` with a defined schema to programmatically pull out clean, usable JSON records, rather than messy text.

### Tracking breaking news on AI policy
A journalist wants the latest updates on EU AI regulations from the last month. They use `chat_with_recency_filter` and restrict results to specific regulatory bodies' websites for maximum relevance.

## Benefits

- Citation Reliability: Instead of accepting general statements, the `chat_with_citations` tool provides a link for every fact, eliminating guesswork and ensuring academic integrity.
- Depth Control: Use `deep_research` when you need an exhaustive report on a complex topic. It takes longer but gives vastly more breadth than quick queries.
- Domain Specificity: The `chat_with_domain_filter` tool lets you narrow searches to trusted domains, which is critical for industry reports or legal research.
- Data Structuring: If your agent needs data for an application, use `structured_query`. It forces the output into reliable JSON format, ready for code consumption.
- Time Sensitivity: When reporting on breaking news, `chat_with_recency_filter` ensures the information is based only on hours or days of old, not outdated knowledge.

## How It Works

The bottom line is you get reliable, verifiable web knowledge delivered directly into your workflow without leaving the chat interface.

1. Subscribe to this MCP and enter your Perplexity API key into your agent's configuration.
2. Instruct your AI client to perform a research query, specifying any necessary filters like domain restrictions or recency requirements.
3. Receive an answer that is not just text, but a synthesized report complete with source links, related questions, and structured data.

## Frequently Asked Questions

**How does chat_with_citations work with Perplexity AI MCP?**
The `chat_with_citations` tool ensures that every single fact or claim in the response is linked back to its original source URL, which is essential for academic integrity.

**Can I restrict my search results using chat_with_domain_filter?**
Yes. You provide a comma-separated list of domains (e.g., 'nih.gov, arxiv.org'), and the agent will only use sources from those trusted websites.

**What is the difference between chat_completion and search_query?**
Use `chat_completion` for a quick answer, but use `search_query` when you want the fullest possible result, which bundles citations, images, and related questions all in one go.

**How do I make sure my research is up-to-date?**
You use `chat_with_recency_filter` and specify the time window (like 'day' or 'week'). This guarantees that your answers are based on recent events, not outdated knowledge.

**Can I ask a follow-up question using chat_with_history?**
Yes. By supplying the conversation history in JSON format, you allow the model to maintain context, so your agent understands exactly what you mean when you ask a follow-up question.