Nativ

Nativ is a macOS app for people who want AI chat and coding tools to use models running on their own Mac. It combines a chat workspace, model library, and local inference server.

• Chat and media: Chat with streaming replies, image attachments, reasoning output, and saved history; compatible models can analyze video, generate or edit images, and transcribe or generate speech.
• Model library: Find installed MLX models and download compatible models from Hugging Face, with hardware recommendations and warnings when models may exceed available memory.
• Coding integrations: Connect Codex, Claude Code, Pi, Hermes, or OpenCode through OpenAI- and Anthropic-compatible APIs, with editor integrations for VS Code, Cursor, and Zed.
• Shared workspace: Keep documents, code, HTML previews, and websites beside chat, where agents can edit shared items or work on the same browser page.
• Performance monitoring: Check response speed, time to first token, memory pressure, and thermal state, with menu bar controls for switching models and managing the server.

Nativ is free and open source under the MIT license. It requires Apple Silicon, macOS 26 or newer, and enough unified memory for your chosen models. MLX-VLM handles inference locally without cloud processing or a required account. Model downloads need internet access, and gated Hugging Face models can require a token. The library includes Gemma 4 E2B, North Mini Code, and LFM2.5-VL 1.6B.