The central place to suggest and track features for Nativ. Pinned so it's easy to find.
How to request
- Search first — this thread and open issues. 👍 an existing request instead of duplicating
- Comment with: what you want, the problem it solves, and roughly how you'd expect it to work.
- One idea per comment keeps voting and discussion clean.
- Found a bug? Please open a separate issue instead.
Status
🟢 Done · 🟡 In progress (PR open / draft / in review) · 🔴 Not started
Maintainers keep this list updated from the thread and open issues:
🟢 Feature 1: Embedding models — serve /v1/embeddings with local embedding models. [#16]
🟡 Feature 2: Let setup continue while model downloads run in the background. [#130]
🔴 Feature 3: Dedicated audio (TTS/STT) support models. [#15]
🔴 Feature 4: Add Core ML / Apple Foundation Models as a backend. [#22]
🟢 Feature 5: Download gated Hugging Face models (auth for license-gated repos). [#23]
🔴 Feature 6: Support macOS Sequoia by lowering the app deployment target. [#24]
🟢 Feature 7: Reduce and document the app bundle footprint. [#26]
🔴 Feature 8: Publish reproducible MLX performance benchmarks by Apple chip. [#27]
🔴 Feature 9: Evaluate DS4 as a backend for DeepSeek V4 Flash. [#28]
🔴 Feature 10: Mac hardware, model, and local-use-case guide. [#29]
🔴 Feature 11: iPad / iPhone version of Nativ. [#30]
🔴 Feature 12: Public roadmap, changelog, and feature support matrix. [#31]
🔴 Feature 13: Expose the model list for editor / VS Code extensions. [#56]
🔴 Feature 14: Guide chat image generation when no image model is selected. [#131]
🔴 Feature 15: Nativ CLI. [#148]
🟢 Feature 16: Homebrew Cask installation. [#57]
🟢 Feature 17: Concurrent model downloads. [#126]
🟢 Feature 18: Sort / organize models by size in the Models page. [#124]
🟡 Feature 19: MCP server support — connect MCP servers so local models can use their tools in Chat. [#154]
The central place to suggest and track features for Nativ. Pinned so it's easy to find.
How to request
Status
🟢 Done · 🟡 In progress (PR open / draft / in review) · 🔴 Not started
Maintainers keep this list updated from the thread and open issues:
🟢 Feature 1: Embedding models — serve
/v1/embeddingswith local embedding models. [#16]🟡 Feature 2: Let setup continue while model downloads run in the background. [#130]
🔴 Feature 3: Dedicated audio (TTS/STT) support models. [#15]
🔴 Feature 4: Add Core ML / Apple Foundation Models as a backend. [#22]
🟢 Feature 5: Download gated Hugging Face models (auth for license-gated repos). [#23]
🔴 Feature 6: Support macOS Sequoia by lowering the app deployment target. [#24]
🟢 Feature 7: Reduce and document the app bundle footprint. [#26]
🔴 Feature 8: Publish reproducible MLX performance benchmarks by Apple chip. [#27]
🔴 Feature 9: Evaluate DS4 as a backend for DeepSeek V4 Flash. [#28]
🔴 Feature 10: Mac hardware, model, and local-use-case guide. [#29]
🔴 Feature 11: iPad / iPhone version of Nativ. [#30]
🔴 Feature 12: Public roadmap, changelog, and feature support matrix. [#31]
🔴 Feature 13: Expose the model list for editor / VS Code extensions. [#56]
🔴 Feature 14: Guide chat image generation when no image model is selected. [#131]
🔴 Feature 15: Nativ CLI. [#148]
🟢 Feature 16: Homebrew Cask installation. [#57]
🟢 Feature 17: Concurrent model downloads. [#126]
🟢 Feature 18: Sort / organize models by size in the Models page. [#124]
🟡 Feature 19: MCP server support — connect MCP servers so local models can use their tools in Chat. [#154]