A pull request submitted to the ggml-org/llama.cpp repository enables local execution of the DeepSeek V4 model.

  • The contribution is authored by user am17an.
  • It is tracked as Pull Request #24162 in the llama.cpp project.