llama.cpp adds support for Aleph-Alpha Kolibri-1 model
October 17, 2026
The llama.cpp repository has merged support for the Kolibri-1 model from Aleph-Alpha. This update enables local inference for the model within the ggml-based runtime architecture.
HOW THIS AFFECTS YOU
●
builderYou can now run Kolibri-1 locally using optimized C++ inference.
●
researcherThis allows for easier benchmarking and testing of the Kolibri-1 architecture on consumer hardware.