Releases of Kimi-Linear-48B-A3B and Step3.5-Flash compatible with llama.cpp are now available.

Details

  • Step3.5-Flash: available at release b7964.
  • Kimi-Linear-48B-A3B: available at release b7957.

Currently, official GGUF files for these models are not yet available on Hugging Face. However, the community is working to make them available.

Ubergarm has already released a GGUF version for Step-3.5-Flash, available on Hugging Face.

The availability of these models in formats compatible with llama.cpp facilitates inference on local hardware, opening new possibilities for those who want to run large language models (LLMs) on-premise.