About Apple RAG MCP
Apple RAG MCP is a documentation retrieval server for Apple platform development. It gives coding agents access to official developer docs and video transcripts, combining keyword and semantic search with Jev relevance ranking.
Key features
- 01Hybrid keyword and semantic search
- 02Jev relevance ranking
- 03Search official Apple documentation
- 04Search WWDC video transcripts
- 05Swift, Objective-C, and SwiftUI examples
Use cases
- 01Find Apple API and platform documentation
- 02Look up technical details from WWDC sessions
- 03Get examples for Swift or SwiftUI development
MCP support
It provides `search` and `fetch` tools to MCP clients over Streamable HTTP, covering Apple documentation and video transcripts.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.apple-rag.comclaude mcp add --transport http apple-rag https://mcp.apple-rag.com{"mcpServers":{"apple-rag":{"type":"http","url":"https://mcp.apple-rag.com"}}}Open in Cursor{"servers":{"apple-rag":{"type":"http","url":"https://mcp.apple-rag.com"}}}Open in VS Code{"mcpServers":{"apple-rag":{"type":"http","url":"https://mcp.apple-rag.com"}}}{"mcpServers":{"apple-rag":{"serverUrl":"https://mcp.apple-rag.com"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://mcp.apple-rag.comAuth requiredTools
2| Parameter | Type | Required | Description |
|---|---|---|---|
| query | string | Yes | Search query for Apple's official developer documentation and video content. Queries must be written in English and focus on technical concepts, APIs, frameworks, features, and version numbers rather than temporal information. |
| result_count | integer | No | Number of results to return (1-10) |
Details
From the project READMEThe Apple docs MCP your AI actually deserves.
Apple docs. WWDC26 transcripts. RAG retrieval + Jev relevance ranking. One clean tool.
· English | 中文
Not Just Another Docs Tool
Hybrid retrieval combines keyword and semantic search to find Apple sources. Jev then ranks the merged candidates for relevance to your question, bringing the best matches to your agent.
Minimal footprint. Maximum signal. Our MCP tools are designed to be lean—no bloated responses, no wasted tokens, no noise cluttering your agent's context. Just the information that matters.
Powered by
Jev
Jev, a decision model, scores candidate Apple documentation and video transcripts against each query's API, platform, and version requirements. We use its relevance scores to rank the sources returned through MCP. A backup reranker takes over if the API call fails.
Open Source. Ready to Use.
The hosted service connects your agent to an Apple knowledge base that is already collected, cleaned, and indexed.
- Ready-to-query sources: Official developer documentation and video transcripts, available through one connection.
- Retrieval model calls included: Embeddings, Jev ranking, and the backup reranker run on our service.
- Ongoing upkeep: We handle content collection, index updates, database backups, and protocol compatibility.
The source is public for inspection and self-hosting. Running your own instance means operating a database, building and refreshing your corpus, and managing your model API credentials.
Start free. Pro includes 50,000 tool calls/week and 50/minute for $1/week, shared by search and fetch.
Start in Seconds
Configure with your agent: In the dashboard, select a token and click Copy Install Prompt. Paste it into your current agent to configure the connection and verify the tools.
One click:
Open Cursor and confirm installation.
Option 2: Manual Setup for Other MCP Clients
JSON Configuration (Copy & Paste): This is a configuration example, not a universal client format. The dashboard also retains manual configuration examples.
{
"mcpServers": {
"apple-rag-mcp": {
"type": "http",
"url": "https://mcp.apple-rag.com"
}
}
}
Manual Configuration Parameters:
- MCP Type:
Streamable HTTP - URL:
https://mcp.apple-rag.com - Protocol:
2026-07-28, with legacyinitializesupport for2025-11-25,2025-06-18, and2025-03-26 - Authentication:
Optional(MCP Token for higher limits) - MCP Token: Get yours at apple-rag.com for increased quota
Saving a configuration alone does not verify a working connection.
Note: No MCP Token required to start! You get free queries without any authentication. Add an MCP Token later for higher usage limits.
🌟 Why Developers Love Apple RAG MCP
⚡ Fast & ReliableGet quick responses with our optimized search infrastructure. No more hunting through docs. 🎯 AI-Powered Hybrid SearchHybrid search combines keyword and semantic retrieval for Apple documentation and video transcripts. Jev then ranks the combined candidates against your question, giving your agent more relevant context. 🔒 Always SecureMCP authentication ensures trusted access for your AI agents with enterprise-grade security. |
📝 Code ExamplesGet practical code examples in Swift, Objective-C, and SwiftUI alongside documentation references. 🔄 Real-time UpdatesOur documentation index is continuously updated for WWDC26, Xcode 27 beta, and the latest Apple developer resources. 🆓 Start FreeTry without an account: 30 tool calls/week and 3/minute. Register at apple-rag.com for 50/week and 5/minute, shared by |
🎯 Features
- 🔍 Semantic Search - Vector similarity with semantic understanding for intelligent retrieval
- 🔎 Keyword Search - Precise technical term matching for API names and specific terminology
- 🎯 Hybrid Search - Combines keyword and semantic retrieval
- 📚 Complete Coverage - iOS 27, iPadOS 27, macOS 27, watchOS 27, tvOS 27, and visionOS 27 documentation
- 🎬 WWDC26 Videos - Full transcripts from Apple Developer videos and WWDC26 sessions
- ⚡ Fast Response - Optimized for speed across all content types
- 🚀 High Performance - Multi-instance cluster deployment for maximum throughput
- 🔄 Always Current - Synced with Apple's latest docs, WWDC26 sessions, and Xcode 27 beta references
- 🛡️ Secure & Private - Your queries stay private
- 🌐 MCP 2026 Native - Stateless MCP
2026-07-28, with legacy Streamable HTTP compatibility
📄 License
This project is licensed under the MIT License.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.
