Nativ: Run frontier open models locally on your Mac(blaizzy.github.io)
363 points by aratahikaru5 1 day ago | 124 comments
tl;dr: Nativ is an open-source (MIT-licensed) macOS app for running open-weight models (from Google, Cohere, Liquid AI) locally on Apple Silicon, built on MLX-VLM. It offers a chat UI with streaming, image input, and live telemetry (tokens/sec, memory, thermals), plus a local endpoint compatible with coding agents like Claude Code, Codex, and OpenCode. The pitch is a fully open alternative to proprietary "local AI" wrappers—no accounts, subscriptions, or cloud dependencies.
HN Discussion:
  • Provides context about the developer's credibility as maintainer of the underlying MLX-VLM library
  • Questions differentiation from existing tools like LM Studio, Jan.ai, and Open WebUI
  • Challenges the 'frontier' framing since true frontier models can't run locally
  • Questions the practical usefulness of small local models for real work
  • ~Shares mixed experiences with MLX reliability compared to llama.cpp