Ollama v0.31.1 Dramatically Speeds Up Gemma 4 on Apple Silicon

Core summary

One sentence to understand this update

Ollama's v0.31.1 release significantly accelerates Gemma 4 model performance on Apple Silicon, achieving nearly 90% faster token generation through multi-token prediction (MTP).

Impact & opportunity

What this could mean

Optimizing local model efficiency is key to improving user experience; builders should continue to track performance enhancements for specific hardware.