Skip to content
mlx.app

Ferret

Apple's research VLM for referring and grounding inside images.

When to use it

Use this when you need region-level grounding — 'what is inside this box' — rather than whole-image captioning or chat.

Alternatives in Vision

MLX-VLM

The go-to package for running vision-language models with MLX.

MobileCLIP

Apple's on-device-sized CLIP for fast image-text matching.

OpenCLIP

The reproducible, community-trained CLIP zoo everyone benchmarks against.