Perplexity
Give your agent live web search, grounded AI answers, and embedding generation so research and docs stay current without manual tab switching.
Overview
com.mcparmory/perplexity is an MCP server for the Idea phase that provides web search, AI-generated answers, and embeddings with up-to-date information.
What is this MCP server?
- Search the web with real-time information through MCP tool calls
- Generate AI responses grounded on retrieved sources
- Create embeddings for RAG pipelines and semantic recall
- stdio transport with uvx PyPI package mcparmory-perplexity 1.0.2
- Optional Docker deploy via ghcr.io/mcparmory/perplexity:1.0.2
- Server version 1.0.2
- 2 registry packages: PyPI and OCI
- Capabilities: search, AI responses, embeddings
Community signal: 25 GitHub stars.
What problem does it solve?
Agents hallucinate or freeze at training cutoff dates when you need current docs, pricing, or competitor pages.
Who is it for?
Builders who want one MCP hook for search-plus-synthesis across early journey phases without maintaining custom scrapers.
Skip if: Production-only teams that forbid external LLM search APIs or need on-prem crawling with no third-party inference.
What do I get? / Deliverables
Your workflow gains live retrieval and embedding tools the agent invokes directly while you discover, validate, or document.
- Search results and synthesized answers in chat
- Embedding vectors for downstream indexes
- Current-source snippets for specs and docs
Recommended MCP Servers
Journey fit
Discover is the primary shelf because Perplexity MCP is most often installed first to widen what the agent can know before scope is fixed, though the same tools support validation and documentation later. Discover captures open-ended exploration—web search and synthesis—rather than a single integration hook tied only to backend code.
How it compares
Live-search and embedding MCP bridge, not a local vector database manager.
Common Questions / FAQ
Who is com.mcparmory/perplexity for?
Solo builders using MCP agents who need web search, grounded answers, and embeddings in the same session as coding.
When should I use com.mcparmory/perplexity?
Use it during discovery and scoping, and again when writing docs or RAG setup that must reflect the current web.
How do I add com.mcparmory/perplexity to my agent?
Register stdio MCP with uvx for mcparmory-perplexity 1.0.2 or the Docker OCI image, and configure your Perplexity API key in the environment your client expects.