Install Wenlan and connect your AI client.

Install the runtime for your system

Wenlan v0.15.8 ships native runtime packages for Windows x64, macOS Apple silicon, and Linux x64 or ARM64 with glibc. Every runtime archive contains the CLI, daemon, and MCP connector.

On Windows, extract wenlan-windows-x64.zip as one unit into a user-owned directory on PATH. Keep onnxruntime.dll and vulkan-1.dll beside the three executables.

# macOS Apple silicon
npx -y wenlan setup

# Linux x64 or ARM64
curl -fsSL https://raw.githubusercontent.com/7xuanlu/wenlan/main/install.sh | bash
wenlan setup --basic
wenlan background on
wenlan status

# Windows x64, after extracting the ZIP and adding it to PATH
wenlan setup --basic
wenlan background on
wenlan status

Claude Code plugin

This is the fastest path. The plugin handles daemon setup, MCP wiring, local memory setup, and the first round-trip check.

If Claude Code asks for a restart after installing, restart once, then run /setup.

/plugin marketplace add 7xuanlu/claude-plugins

/plugin install wenlan@7xuanlu

/setup

Codex

Run Wenlan setup, then connect Codex to the local MCP server. The main Wenlan repository also ships a Codex plugin for users installing from a checkout; wenlan connect codex is the direct no-checkout client path.

wenlan connect codex

ChatGPT and Claude.ai

Open Remote Access in the Wenlan desktop app to create a temporary HTTPS URL for Streamable HTTP MCP. The app starts wenlan-mcp with --no-auth on loopback and exposes it through the tunnel, so possession of the URL grants access. In ChatGPT, open Settings > Plugins, create a New Plugin, choose Server URL under Connection, paste the URL, and set Authentication to None. In Claude.ai, install Wenlan from the 7xuanlu/wenlan marketplace through Directory > Plugins.

This is a custom MCP connection to your own Wenlan runtime, not a claim that Wenlan is listed in the public ChatGPT Apps Directory. Stop Remote Access when you are not using it.

Other local MCP clients

For Cursor, Claude Desktop, Gemini CLI, VS Code, and other supported local MCP clients, set up the Wenlan runtime first. Then let the CLI write the client-specific MCP configuration.

Wenlan setup installs the CLI, daemon, and MCP connector, registers the daemon with your operating system's user service manager, and verifies status.

wenlan connect cursor
# or: claude-desktop, vscode, gemini

What to try first

Store one durable project fact, then ask another session or client to recall it. Wenlan should surface the fact and keep its source available to the wiki and review flow.