About Edlide

Edlide is an AI-powered IDE focused on open source models and privacy. Unlike other AI IDEs, Edlide does not collect or retain your important data, ensuring a secure and private development environment.

Open Source Focus

Work with the best open source AI models without vendor lock-in.

Privacy First

No data retention - your data stays yours.

Installation

Get started by downloading Edlide for your platform.

Go to Download Page

Download the appropriate installer for macOS or Windows.

macOS

Download the DMG file and install Edlide on your Mac.

Windows

Download the installer and install Edlide on Windows.

Account Setup

Connect your Edlide IDE to your account to access all features.

Open Settings

Navigate to Settings → Account in the IDE.

Connect Account

Click Connect to your account to open the Edlide website.

Authenticate

Transfer Settings

Import your settings from other IDEs seamlessly.

Available transfer options in Settings → General:

• Transfer from VS Code - Import extensions and settings
• Transfer from Cursor - Migrate your Cursor configuration
• Transfer from Windsurf - Import Windsurf settings

Edlide CLI

A terminal-based AI coding agent. Same models, same account, works anywhere.

1. Install

Download and install the CLI from the download page:

Go to Download Page

Find the CLI install command in the CLI section.

2. Connect your account

After installation, authenticate with your Edlide account:

/connect

This opens a browser window where you sign in. The CLI links to your account automatically.

3. Start coding

Launch the agent in any project directory. Works the same way as Edlide IDE. Same models, same requests quota.

edlide

MCP Support

Supports the MCP format. Same server configurations as in the IDE.

.edliderules

Project-level rules file. Place in your project root. The CLI reads it automatically.

EDLIDE.md

Project memory file. The agent reads and updates it to maintain context across sessions.

Available Models

Choose from the best open source AI models.

Model	Context Window
minimax-m2.5	200k
glm-4.7	200k
kimi-k2.5	256k

Key Features

No Context Window Limits

Work with large contexts without artificial limitations.

Native Quantization

Optimized for performance and efficiency.

System Prompt

Configure your global communication style and AI behavior in Settings → Rules → System Prompt.

The System Prompt applies globally and does not change when opening different projects. Use it to define communication style, explanation depth, and other general AI behavior preferences.

Project Rules

Create project-specific rules for tailored AI responses.

Create Rule

Click the + button in Settings → Rules.

Name Your Rule

Enter a name for your project-specific rule.

File Created

A file is automatically created in your project with a .edliderules extension.

Project rules are ideal for specific project requirements that need tailored AI behavior for better results. Each rule applies only to its respective project.

MCP Setup

Model Context Protocol (MCP) servers extend Edlide capabilities. First, ensure Node.js is installed.

Open MCP Settings

Navigate to Settings → MCP.

Add Server

Click Add MCP Server to configure your server.

Example Configuration

Add MCP servers by pasting this configuration into your MCP settings:

{
  "mcpServers": {
    "supabase": {
      "command": "npx",
      "args": [
        "-y",
        "@supabase/mcp-server-supabase@latest",
        "--access-token",
        "YOUR_SUPABASE_ACCESS_TOKEN"
      ],
      "transportType": "stdio"
    },
    "search": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-brave-search"
      ],
      "env": {
        "SEARCH_API_KEY": "YOUR_API_KEY"
      }
    },
    "context": {
      "command": "npx",
      "args": [
        "-y",
        "@upstash/context7-mcp"
      ]
    },
    "thinking": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-sequential-thinking"
      ]
    },
    "docs": {
      "url": "https://example.com/mcp-docs"
    }
  }
}

Apply Settings

Configure how the Apply button behaves in Settings → Actions → Apply.

Same as Chat Model

Use the same AI model for Apply functionality as your chat model.

Tools

Functions that LLMs can call. Some tools require user approval before execution.

Auto-Approve Settings

Configure automatic approval behaviors in Settings → Actions.

Auto-approve edits

Automatically accepts code edits suggested by the AI without requiring manual confirmation.

Auto-approve terminal

Automatically accepts terminal command executions run by the AI.

Auto-approve MCP tools

Automatically accepts MCP (Model Context Protocol) tool calls made by the AI.

Fix lint errors

Automatically fixes lint errors detected in your code without asking for confirmation.

Auto-accept LLM changes

Automatically accepts all changes made by the LLM during conversations and edits.

Autocompacting

Edlide automatically manages context window efficiency when it reaches 80% capacity.

Threshold Reached

When context window reaches 80%, the auto-compacting mechanism triggers.

Summarization

The entire chat session is summarized to preserve key information.

New Session

A fresh session opens with the summarized context for better performance.

Context Window

Monitor your real-time context window usage in the chat interface.

Hover over the context window indicator in the chat for 2+ seconds to see your real-time context usage. This helps you understand when auto-compacting will be triggered.