Dive
Dive is an open-source MCP Host Desktop Application that seamlessly integrates with any LLMs supporting function calling capabilities. ✨
What is Dive?
Dive is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to dive is an open-source mcp host desktop application that seamlessly integrates with any llms supporting function calling capabilities. ✨
Dive is an open-source MCP Host Desktop Application that seamlessly integrates with any LLMs supporting function calling capabilities. ✨
This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Dive is an open-source MCP Host Desktop Application that sea
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx diveConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Dive
Dive is an open-source desktop application that acts as an MCP Host, letting users connect any LLM with function-calling support—including Claude, ChatGPT, Gemini, and local Ollama models—to any MCP server for agent-like workflows. It runs on Windows, macOS, and Linux, requires no cloud subscription for local model use, and provides a unified chat interface with built-in tools for file management, web requests, and shell execution. Engineers and power users choose Dive when they want a single cross-platform front-end for mixing and matching AI models with MCP tool servers.
Prerequisites
- Windows 10+, macOS 12+, or a modern Linux distribution (Ubuntu 22.04+ recommended)
- Node.js 18+ and Python 3.10+ (auto-downloaded on Windows launch; manual install required on macOS/Linux)
- API key for at least one provider: Anthropic (Claude), OpenAI (ChatGPT), or Google (Gemini); or Ollama installed locally for offline use
- An internet connection for cloud model providers or Ollama running locally for offline inference
Download and install Dive
Go to the GitHub Releases page at github.com/OpenAgentPlatform/Dive/releases/latest and download the installer for your platform. On Windows use the Tauri (.exe) build for the smallest footprint; on macOS use the .dmg; on Linux use the .AppImage or install via paru on Arch.
# Arch Linux only:
paru -S dive-aiLaunch Dive and configure a model provider
Open Dive and navigate to Settings → Model. Enter your API key for the desired provider (Anthropic, OpenAI, Google, or an Ollama-compatible endpoint) and select a model to use as the default.
Add an MCP server
In Settings → MCP Servers, click Add Server. Paste the server's stdio command or HTTP/SSE endpoint URL. Dive will auto-start stdio servers and connect to remote HTTP servers.
Enable or disable individual tools
After adding an MCP server, Dive lists every tool it exposes. Toggle individual tools on or off per session to control what the model can invoke, reducing noise and improving response quality.
Start a chat session
Return to the main Chat view. Dive injects all enabled MCP tool definitions into the model context automatically. Type a natural-language request and the model will call tools as needed.
Use built-in tools without any MCP server
Dive ships with built-in tools for web fetch, file read/write/list, and bash execution. These are available in every session even without configuring a separate MCP server.
Dive Examples
MCP server configuration in Dive
Dive stores MCP server configurations in a JSON file. This example shows adding the filesystem and fetch servers.
{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/projects"]
},
"fetch": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-fetch"]
}
}
}Prompts to try
With the built-in tools and an MCP server configured, try these prompts in the Dive chat interface.
- "List all files in my projects folder and summarize the largest ones"
- "Search the web for the latest Claude API pricing and save a summary to notes.md"
- "Run npm test in /Users/you/myapp and show me any failures"
- "Fetch https://example.com and extract all external links"Troubleshooting Dive
Dive fails to start on Linux with a sandbox error
Launch the AppImage with the --no-sandbox flag: ./Dive.AppImage --no-sandbox. This is required on some Ubuntu/Debian systems due to kernel namespace restrictions.
Ollama models do not appear in the model selector
Ensure Ollama is running (ollama serve) and set the base URL to http://localhost:11434 in Dive Settings → Model → Custom Endpoint. Pull a model first with ollama pull llama3 before selecting it.
MCP server tools do not appear in chat
Check that the server command is correct and the binary is on PATH. Open the MCP Servers settings page and look for a connection error indicator. For stdio servers, try running the command manually in a terminal to confirm it starts without errors.
Frequently Asked Questions about Dive
What is Dive?
Dive is a Model Context Protocol (MCP) server that dive is an open-source mcp host desktop application that seamlessly integrates with any llms supporting function calling capabilities. ✨ It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Dive?
Follow the installation instructions on the Dive GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Dive?
Dive works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Dive free to use?
Yes, Dive is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Dive Alternatives — Similar Developer Tools Servers
Looking for alternatives to Dive? Here are other popular developer tools servers you can use with Claude, Cursor, and VS Code.
Ecc
★ 188.2kThe agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Javaguide
★ 155.8kJava 面试 & 后端通用面试指南，覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发
Gemini CLI
★ 104.5kA secure MCP server that wraps the Google Gemini CLI, allowing clients to query Gemini models using local OAuth sessions without requiring an API key. It provides tools for model interaction and diagnostics with built-in protection against command in
Awesome MCP Servers
★ 87.3k⭐ Curated list of Model Context Protocol (MCP) servers - tools that extend Claude Desktop, Cursor, Windsurf, and other MCP clients with custom capabilities.
MCP Servers
★ 86.0kModel Context Protocol Servers
CC Switch
★ 77.5kA cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes Agent. Only official website: ccswitch.io
Browse More Developer Tools MCP Servers
Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Dive in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Dive?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.