Ollama v0.31.1: Faster Gemma 4 on Apple Silicon

Core summary

One sentence to understand this update

Ollama v0.31.1 dramatically improves Gemma 4's performance on Apple Silicon, achieving nearly 90% faster token generation through multi-token prediction.

Impact & opportunity

What this could mean

Developers building AI applications on Apple Silicon with Ollama can now experience significant speedups when using Gemma 4, enhancing efficiency for coding agents and other tasks.