Overview

The MCP Redis Server enables AI models to interact with Redis, the world's most popular in-memory data structure store. Redis is widely used for caching, real-time analytics, session management, and message queuing, making it essential for high-performance applications.

Official Server:

Developed and maintained by Anthropic

Key Features

⚡

High-Performance Operations

Lightning-fast key-value operations with in-memory data storage

🔑

Flexible Key Management

Set, get, delete, and search keys with pattern matching support

⏰

TTL & Expiration

Automatic key expiration with Time-To-Live configuration

💾

Caching & Sessions

Ideal for caching, session storage, and rate limiting

Available Tools

Quick Reference

Tool	Purpose	Category
`redis_get`	Retrieve value by key	Read
`redis_set`	Store key-value pair	Write
`redis_delete`	Remove one or more keys	Write
`redis_list`	List keys by pattern	Discovery

Detailed Usage

redis_get▶

Retrieve the value associated with a Redis key.

use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_get",
  arguments: {
    key: "user:1234:profile"
  }
});

Returns the stored value or null if key doesn't exist.

redis_set▶

Store a key-value pair in Redis with optional expiration time.

// Set a simple key-value pair
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: "session:abc123",
    value: JSON.stringify({ userId: 1234, role: "admin" })
  }
});

// Set with 1-hour expiration (TTL)
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: "cache:trending",
    value: JSON.stringify({ items: [...] }),
    expireSeconds: 3600
  }
});

TTL automatically removes keys after expiration.

redis_delete▶

Delete one or more keys from Redis.

// Delete a single key
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_delete",
  arguments: {
    key: "temp:data"
  }
});

// Delete multiple keys at once
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_delete",
  arguments: {
    key: ["cache:page1", "cache:page2", "cache:page3"]
  }
});

Returns the number of keys successfully deleted.

redis_list▶

List all keys matching a pattern using Redis glob-style patterns.

// List all keys
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_list",
  arguments: {
    pattern: "*"
  }
});

// List keys with specific prefix
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_list",
  arguments: {
    pattern: "user:*"
  }
});

// List session keys
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_list",
  arguments: {
    pattern: "session:*"
  }
});

Supports glob patterns: * (any chars), ? (single char), [abc] (char set)

Installation

{
  "mcpServers": {
    "redis": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-redis",
        "redis://localhost:6379"
      ]
    }
  }
}

Custom Connection:

Replace redis://localhost:6379 with your Redis connection URL. For password-protected instances, use redis://:password@host:port

Common Use Cases

1. Session Management

Store user sessions with automatic expiration:

// Store session with 24-hour expiration
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: "session:user123",
    value: JSON.stringify({
      userId: 123,
      loginTime: Date.now(),
      permissions: ["read", "write"]
    }),
    expireSeconds: 86400  // 24 hours
  }
});

2. Caching API Responses

Cache expensive API calls to improve performance:

// Cache API response for 5 minutes
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: "api:weather:london",
    value: JSON.stringify(weatherData),
    expireSeconds: 300
  }
});

3. Rate Limiting

Track API request counts per user:

// Increment request counter
const key = `ratelimit:${userId}:${currentMinute}`;
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: key,
    value: requestCount.toString(),
    expireSeconds: 60
  }
});

4. Temporary Data Storage

Store temporary processing results:

// Store job results for 1 hour
use_mcp_tool({
  server_name: "redis",
  tool_name: "redis_set",
  arguments: {
    key: `job:${jobId}:result`,
    value: JSON.stringify(processingResult),
    expireSeconds: 3600
  }
});

Connection String Format

The Redis MCP server accepts standard Redis connection URLs:

Local: redis://localhost:6379
With password: redis://:password@host:6379
Custom database: redis://host:6379/2
TLS/SSL: rediss://host:6379
With auth & DB: redis://:password@host:6379/1

Sources

PostgreSQL MCP Server

PostgreSQL MCP servers enable AI models to interact with PostgreSQL databases, providing capabilities for structured data operations, SQL queries, transaction management, and advanced data indexing.

Supavec MCP Server: Vector Database for AI Applications

Supavec MCP servers enable AI models to interact with vector databases, providing capabilities for storing, searching, and managing vector embeddings for AI applications.

Supabase MCP Server

Supabase MCP servers enable AI models to interact with Supabase backends, providing capabilities for PostgreSQL databases, real-time subscriptions, authentication, and storage operations.

Redis MCP Server