Ollama v0.31.1 Speeds Up Gemma 4 on Apple Silicon by 90%

Core summary

One sentence to understand this update

Ollama v0.31.1 dramatically accelerates Gemma 4 performance on Apple Silicon, achieving nearly 90% faster token generation using multi-token prediction.

Impact & opportunity

What this could mean

Apple Silicon users will experience significantly faster and more efficient local Gemma 4 model inference, benefiting local development and testing workflows.