Ollama v0.31.1 Accelerates Gemma 4 on Apple Silicon

Core summary

One sentence to understand this update

Ollama v0.31.1 release significantly boosts Gemma 4 performance on Apple Silicon, achieving nearly 90% faster token generation through multi-token prediction.

Impact & opportunity

What this could mean

Developers using Ollama on Apple Silicon can now experience much faster inference with Gemma 4, improving local AI development and deployment efficiency.