2026
03
24
22
38
[AI] build llama.cpp in windows
Before that, you need to install NVIDIA's CUDA Toolkit in windows system.
open "Developer Command Prompt for VS 2022"
>%comspec% /k "C:\Program Files\Microsoft Visual Studio\2022\Community\Common7\Tools\VsDevCmd.bat
goto vcpkg directory
> vcpkg install curl
> vcpkg install openssl
open "Developer Command Prompt for VS 2022"
>%comspec% /k "C:\Program Files\Microsoft Visual Studio\2022\Community\Common7\Tools\VsDevCmd.bat
goto vcpkg directory
> vcpkg install curl
> vcpkg install openssl
goto llama.cpp directory
> mkdir build
> cd build
> cmake .. -DGGML_CUDA=ON -DCMAKE_TOOLCHAIN_FILE=[vcpkg path]/scripts/buildsystems/vcpkg.cmake
> cmake --build . --config Release -n 2
PS: The llama.cpp Git repository may occasionally get stuck on specific commits. Please ensure you are using the latest version.