Ollama v0.30.8 Released with Performance and Stability Improvements

Core summary

One sentence to understand this update

Ollama released v0.30.8, which fixes issues with incorrect provider selection, improves prompt caching by decoupling it from context shift for better KV cache reuse, and offers more stable MLX inference.

Impact & opportunity

What this could mean

Ollama users can expect a more stable and performant model inference experience, particularly with improved MLX inference and caching efficiency.