// event · tool
Ollama v0.40.0 runs models on MLX by default on Apple Silicon
toolvendor: ollamapublished: 18d agoconfidence: 0.85id: f134ddab-85d9-4871-aa0c-6710bed86cb4
Ollama 0.40.0 automatically runs supported model architectures on Apple's MLX runtime on Apple Silicon, delivering better native performance. New MLX-enabled models include gemma4, qwen3.5/3.6/3.8, decision models (Nimble, tev1, clef, clef-flash), and the embeddinggemma-2 embedding model.
capability delta
An agent can now run local LLMs on Apple Silicon through Ollama with MLX acceleration by default, including new gemma4, qwen3.x, decision models, and an MLX embedding model, without extra configuration.
curator note
Significant runtime update: Ollama now defaults to MLX on Apple Silicon, which materially improves local inference performance for agents running on Mac hardware.
// install
brew install ollamagithub · ollama/ollama @ v0.40.0
// example
bash
# Pull and run a model with MLX acceleration on Apple Silicon
ollama pull qwen3.8
ollama run qwen3.8
# Use the new MLX embedding model
ollama pull embeddinggemma-2
curl http://localhost:11434/api/embeddings -d '{
"model": "embeddinggemma-2",
"prompt": "agents need fast local embeddings"
}'// links