L
Localthink
Verified IntegrationAuthor: @H3xabah•Category: ServerApplication
JSON-RPC 2.0
Protocol Standard
Sub-second
Execution Latency
Active
Operational Status
Client Configuration
— Connect Localthink to Claude Desktop or Cursor in seconds{
"mcpServers": {
"localthink": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-localthink"
],
"env": {}
}
}
}Paste into
~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).Architecture & Capabilities
System Overview
Offload large file queries to a local LLM for Claude Code, preserving context and significantly reducing token usage.
Indexed Date
7/23/2026
License
Open Source
Protocol Layer
stdio / SSE RPC
Frequently Asked Questions
Architecture and operational details for Localthink
You need Ollama installed and running, the Claude Code CLI, and Python 3.10+. Installation is straightforward using `claude mcp add localthink`.
Related MCP Servers
Browse all servers →C
@carboneio
CarboneEmpower AI assistants to generate and convert documents, manage templates, and automate document workflows efficiently.
ClaudeSearchDatabase+65 FAQs
C
@sebastienrousseau
CloudCDNProvides a multi-tenant, AI-native Content Delivery Network deployable on Cloudflare, featuring sub-100ms TTFB, AI agent controllability, and comprehensive accessibility.
ClaudeSearchDatabase+65 FAQs
C
@sargioboys51
Cloudflare AI Agent ServerDeploy a Model Context Protocol server on Cloudflare Workers without requiring authentication.
ClaudeSearchDatabase+55 FAQs