Brave Search MCP Server vs Perplexity MCP Server
Two ways to give Claude live web knowledge: Brave Search returns classic search results via the Brave API, keeping Claude in control of reading and synthesis, while the Perplexity server delegates the question to Perplexity's own answer engine. Brave suits agents that want raw results to reason over; Perplexity suits pipelines that want a synthesized answer in one hop. Compare their maintenance activity below — it differs more than their taglines do.
GitHub data verified 2026-07-30 — objective fields only, no editorial scoring.
| Brave Search MCP Server Web search powered by Brave's privacy-focused search engine | Perplexity MCP Server AI-powered research and search using Perplexity API | |
|---|---|---|
| GitHub stars | 1,335 | 306 |
| Forks | 189 | 47 |
| Open issues | 30 | 8 |
| Last commit | 2026-07-28 | 2025-03-27 |
| Maintenance status | Actively maintained | Actively maintained |
| License | MIT | MIT |
| Language | TypeScript | TypeScript |
| Install | npx @modelcontextprotocol/server-brave-search | — |
| Author | Anthropic | Perplexity AI |
| Official listing | Yes | No |
| Community votes | 0 | 0 |
| Copies from directory | 0 | 0 |
GitHub statistics are measured at repository level; servers that live in a shared monorepo show that repository's stats. This comparison is generated from directory data — no sponsored placement, see the curation policy.