Download Wenlan
Choose one published build, keep its files together, and verify the local runtime before connecting an AI tool.
Choose your build
The current release provides four prebuilt desktop archives.
We could not identify a supported desktop build in this browser.
Windows
x64
A complete native archive with ONNX Runtime and the Vulkan loader for supported GPUs. No WSL or Rust toolchain required.
ZIP · 66.3 MB
CLI · daemon · MCP connector
macOS
Apple silicon
Headless runtime with Metal support for local model paths.
TAR.GZ · 45.6 MB
CLI · daemon · MCP connector
Linux
x64 · glibc
Prebuilt runtime for common x86_64 Linux systems.
TAR.GZ · 56.5 MB
CLI · daemon · MCP connector
Linux
ARM64 · glibc
Prebuilt runtime for aarch64 Linux systems.
TAR.GZ · 50.7 MB
CLI · daemon · MCP connector
Install on your platform
Each archive contains the CLI, local daemon, and MCP connector.
Windowsx64
- 01Download and extract the ZIP. Keep every included file together.
- 02Add the extracted directory to PATH.
- 03Open a new terminal and run wenlan doctor.
macOSApple silicon
- 01Download and extract the TAR.GZ archive.
- 02Move the binaries to a directory on PATH.
- 03Run wenlan doctor to verify the local runtime.
Linuxx64 · glibc
- 01Download and extract the TAR.GZ archive.
- 02Move the binaries to a directory on PATH.
- 03Run wenlan doctor to verify the local runtime.
LinuxARM64 · glibc
- 01Download and extract the TAR.GZ archive.
- 02Move the binaries to a directory on PATH.
- 03Run wenlan doctor to verify the local runtime.
Verify before you connect
Run the diagnostic after installation. It checks the local runtime and reports the next repair step.
wenlan doctor