Perplexity API Platform

Official

Real-time search, reasoning, and research through Perplexity's API.

How to install ↓ Compare GitHub source repository ↗

Category: Other #42 of 230
Stars: ★ 2.6k Very popular
Transport: stdio (local process)
Runtime: Node.js · Docker
Credentials: API key / credential required
License: MIT
Last commit: 10 days ago
Tools: 4

64FMRS · C

The official Perplexity MCP Server provides four tools for real-time search, conversational answers, deep research, and advanced reasoning. It requires a Perplexity API key and network access, and uses stdio by default.

Strongest · Documentation 15/20 Weakest · Reliability 10/20

Reliability

10/20

Security and permissions

12/20

Maintenance

13/20

Documentation

15/20

Setup experience

14/20

Why each score

Reliability 10/20: The manifest declares an npm stdio package at version 1.2.0, and the README identifies four tools with basic behavior, making the normal initialization path plausible. However, the supplied material contains no verifiable source, CI, or committed tests covering MCP initialization, tool-list handshake, error handling, timeouts, or dependency behavior. This warrants a conservative deduction and keeps the score at or below 12 under static calibration.
Security and permissions 12/20: The API key is supplied through an environment variable, and no real secret appears in the examples. HTTP mode defaults to loopback and documents host and origin restrictions, showing some attention to boundaries. Points are deducted because the server performs external search/API requests without clear disclosure of query-data boundaries, privacy handling, cost controls, or user confirmation; custom base URLs and proxies also expand the network trust boundary. No evidence of a security red line was provided.
Maintenance 13/20: The repository is marked unarchived, uses the MIT license, and provides a recognizable official package and explicit version, indicating identifiable ownership and a usable update path. The supplied material does not verify commit or release cadence, dependency maintenance, issue response, vulnerability disclosure, or a security-response channel. Stars and open-issue counts are discovery signals and do not add points.
Documentation 15/20: The README covers installation, API-key configuration, the four tool purposes, multiple client examples, proxy handling, HTTP deployment, library usage, and troubleshooting. This is substantially usable documentation. Points are deducted because reviewable tool schemas, return formats, rate/context/timeout limits, cost information, and data-permission disclosures are missing; the evidence is primarily the untrusted README, without source or test cross-checking.
Setup experience 14/20: The project offers a simple npx installation command, stdio configuration, examples for several mainstream clients, and guidance for credentials and common failures. Points are deducted because Node/npm and a valid API key are required, an npx stdout issue may require an extra adjustment, and HTTP/proxy deployment adds complexity. No CI or committed tests verify installation or the MCP handshake, so the static calibration limit is respected.

Static review · not runListed 2026-07-29

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

AI assistants that need real-time web search
Users needing Perplexity Agent API research and reasoning capabilities
Developers using MCP-compatible client tools

Not for

Local tasks that do not need internet or an external API
Environments without a Perplexity API key
Immediate-use scenarios that cannot wait for research runs lasting several minutes

Required permissions

Requires access to PERPLEXITY_API_KEY
Requires network access to the Perplexity API and the internet
HTTP deployments may require network binding and host, origin, and port configuration

Risks and side effects

The API key must be configured through the environment and should not be exposed.
Setting BIND_ADDRESS to 0.0.0.0 exposes the HTTP service on all network interfaces.
Setting ALLOWED_ORIGINS to * permits cross-origin browser requests from any origin.
Deep research runs may take minutes and may require a higher timeout.

Setup

Before you start

Runtime：Node.js · Docker

PERPLEXITY_API_KEY requiredsecret Required Perplexity API key, obtained from the Perplexity API Portal (perplexity.ai/account/api/group)

Other optional settings (10)

PERPLEXITY_TIMEOUT_MS optional Optional API request timeout in milliseconds, default 5 minutes; raise it for long research runs

PERPLEXITY_BASE_URL optional Optional custom API base URL, default https://api.perplexity.ai

PERPLEXITY_LOG_LEVEL optional Optional log level: DEBUG|INFO|WARN|ERROR, default ERROR

PERPLEXITY_PROXY optional Optional corporate proxy URL (must include https://), supports username:password form

HTTPS_PROXY optional Optional standard HTTPS proxy variable, lower priority than PERPLEXITY_PROXY

HTTP_PROXY optional Optional standard HTTP proxy variable, lowest priority

PORT optional HTTP-mode server port, default 8080

BIND_ADDRESS optional HTTP-mode network interface to bind to, default 127.0.0.1; set to 0.0.0.0 to expose on all interfaces

ALLOWED_ORIGINS optional HTTP-mode CORS origins (comma-separated), default empty (no cross-origin browser requests)

ALLOWED_HOSTS optional HTTP-mode additional allowed Host header values (comma-separated); add the public hostname when binding to 0.0.0.0

Get an API key from the Perplexity API Portal, set PERPLEXITY_API_KEY to that key, and run npx -y @perplexity-ai/mcp-server using the supplied client configuration. For long research tasks, set PERPLEXITY_TIMEOUT_MS. Corporate networks may require PERPLEXITY_PROXY, HTTPS_PROXY, or HTTP_PROXY.

.mcp.json

{"mcpServers":{"perplexity":{"command":"npx","args":["-y","@perplexity-ai/mcp-server"],"env":{"PERPLEXITY_API_KEY":"your_key_here"}}}}

Shown for Claude Code. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json

{
  "servers": {
    "perplexity": {
      "command": "npx",
      "args": [
        "-y",
        "@perplexity-ai/mcp-server"
      ],
      "env": {
        "PERPLEXITY_API_KEY": "your_key_here"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal

claude mcp add perplexity -e PERPLEXITY_API_KEY=your_key_here -- npx -y @perplexity-ai/mcp-server

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

After configuring PERPLEXITY_API_KEY and starting the server, the client's tool list should show perplexity_search, perplexity_ask, perplexity_research, and perplexity_reason; calling perplexity_ask with a current-events question that returns sourced results confirms the connection works.

Troubleshooting

Verify that PERPLEXITY_API_KEY is set correctly and valid.
Check network connectivity and whether a firewall or proxy allows access to api.perplexity.ai.
Confirm that the npm package is installed and the command path is correct.
Increase PERPLEXITY_TIMEOUT_MS for long research queries.
For EOF or Initialize errors, use npx -yq to suppress installation output.

Things to try

Once connected, you can ask your AI assistant things like:

Use perplexity_search to find the latest AI news from the past week only
Use perplexity_ask to get the current Bitcoin price
Use perplexity_research to write a detailed report on the current state of solid-state battery technology
Use perplexity_reason to analyze why my distributed system is showing consistency issues

Tools 4

perplexity_search read-only

Direct web search through the Perplexity Search API, returning ranked results with metadata and supporting recency and domain filters.

perplexity_ask read-only

General-purpose conversational AI backed by the Agent API fast preset, with real-time web search for quick questions and everyday searches.

perplexity_research read-only

Deep research backed by the Agent API high preset for comprehensive analysis and detailed reports; runs may take minutes.

perplexity_reason read-only

Advanced reasoning and problem-solving backed by the Agent API medium preset for complex analytical tasks.

Use cases

Finding current information on the web

Answering quick questions and performing everyday searches

Producing comprehensive research reports

Handling complex analysis and reasoning tasks

Supported clients

Claude Code

Codex

Cursor

Claude Desktop

Kiro

Windsurf

VS Code

Listed from the project's documentation, not tested by this site.

Overview

The official MCP server implementation for the Perplexity API Platform. It gives AI assistants real-time web search, reasoning, and research capabilities through the Agent API and Search API. The server runs over stdio, and the npm package is @perplexity-ai/mcp-server 1.2.0.

Similar servers

Hevy MCP 77 · B

Manage your Hevy workout data from AI assistants via MCP.

★ 517 · Tools 26 Compare with this →

Firecrawl MCP Server 75 · B

Firecrawl's official MCP server for web search, scraping, and structured extraction for AI agents

★ 7.6k · Tools 11 Compare with this →

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →

Compartment 75 · B

Encrypted, fully offline long-term memory for AI agents

★ 583 · Tools 14 Compare with this →

Compare with other MCP servers

Source revision 6ba09f856ab8 Data synced 2026-10-19 Read the FMRS scoring method