Ollama v0.40.12 released with updates to MLX and llama.cpp
Key point
Ollama v0.40.12 has been released, including improvements to preserve small-image detail for Gemma 4 vision towers in MLX and a llama.cpp version bump.
Ollama v0.40.12 has been released. Regarding MLX, updates include preserving small-image detail for compact Gemma 4 vision towers, a version bump for MLX and MLX-C, and prompt speedups for gemma/flint/laguna. Additionally, DFlash2 support has been added to MLX.
For llama.cpp, the version has been bumped to b11901. This release includes v0.40.11 and v0.40.12-rc0.
This item is a summary. The full, checkable version is at the original source:
Read the original at Ollama ↗