v0.3.5

What's Changed

• Allow slow local models to start by @Lazarus-931 in #380
• Revert "Feature/live server settings" by @Blaizzy in #383
• Improving file processing by @Lazarus-931 in #375
• Keep models whose shard index describes another build by @lizen in #370
• Simplify custom MCP stdio setup by @Lazarus-931 in #388
• Add an optional tool to read files safely from a user-selected folder by @lucasnewman in #389
• Document Homebrew installation path by @jacquerie in #392
• Stabilize Kit definitions and additive activation by @Lazarus-931 in #394
• Add GLM-5.3-Flash (Z.ai) to partner models by @Lazarus-931 in #398
• Resolve Kit capabilities at routine runtime by @Lazarus-931 in #395
• Bump version to 0.3.5 by @lucasnewman in #399
• Discover third-party server tools installed by Homebrew by @Lazarus-931 in #391

New Contributors

• @lizen made their first contribution in #370
• @jacquerie made their first contribution in #392

Full Changelog: v0.3.4...v0.3.5

v0.3.4

What's Changed

• Report temperature, fans, and power to the system stats tool by @Lazarus-931 in #328
• Improve gated model download errors by @Lazarus-931 in #333
• Don't block the main thread when loading the stats database by @lucasnewman in #342
• Reconcile routine launchctl jobs off the main thread to avoid blocking app startup by @lucasnewman in #341
• Menu handling cleanup by @lucasnewman in #340
• Fix multi-key audio shortcut capture by @Lazarus-931 in #348
• Fix non-disappearing button when creating a schedule by @Lazarus-931 in #345
• missing enter support for popups by @Lazarus-931 in #344
• Check for software updates only after app launch is complete by @lucasnewman in #343
• improve xet download progress bar UI by @Yavaren in #316
• Control panel cleanup by @lucasnewman in #351
• Make live resizing of the sidebar smooth and avoid truncating titles unnecessarily by @lucasnewman in #352
• Rework control panel for smooth animations and consistent layout by @lucasnewman in #353
• Move server logs into a dedicated observable store by @Lazarus-931 in #357
• Surface chat persistence write failures by @Lazarus-931 in #355
• Improve server termination logging by @Lazarus-931 in #350
• Store the Hugging Face token in the Keychain by @Lazarus-931 in #362
• Migrate SystemMonitorStore to @observable by @Lazarus-931 in #356
• Migrate NativModel to @observable by @Lazarus-931 in #359
• Move audio meter capture off the main thread by @lucasnewman in #364
• Feature/live server settings by @Lazarus-931 in #315
• Surface MCP connection failures by @Lazarus-931 in #367
• Add audio upload to the library by @Blaizzy in #331
• Bump version to 0.3.4 by @lucasnewman in #368
• Fix scrolling performance for long user prompts by @lucasnewman in #373
• Decouple scheduled tasks from chats by @Blaizzy in #371
• Fix the scroll indicator overlapping with the composer gradient by @lucasnewman in #374
• Fix the material overlay in dark mode by @lucasnewman in #377
• Adjust detail header padding for sidebar/fullscreen states by @lucasnewman in #378

Full Changelog: v0.3.3...v0.3.4

v0.3.3

What's Changed

• fix core audio callback isolation by @Yavaren in #320
• Bump version to 0.3.3 by @Blaizzy in #326
• Fix audio transcription configuration lifetime by @Yavaren in #322
• Add dedicated scheduled tasks workspace by @Yavaren in #284

Full Changelog: v0.3.2...v0.3.3

v0.3.2

What's Changed

• Polish sidebar and model controls by @Blaizzy in #293
• Recover from a frozen chat instead of locking the UI by @Lazarus-931 in #286
• Add reranking capability discovery by @TheAngryPit in #302
• Revert the streaming markdown rendering by @Lazarus-931 in #294
• Fix microphone permission request flow by @Yavaren in #292
• Bump version to 0.3.2 by @lucasnewman in #305
• Restrict the state observed by the sidebar to reduce steady state relayout by @lucasnewman in #307
• Render fully-styled streamed chat content incrementally by @lucasnewman in #308
• Working on global MCP tool skill kit and extent and extension management by @Lazarus-931 in #281
• Update language mode to Swift 6.3 and fix all concurrency warnings by @lucasnewman in #311
• Move ChatViewModel to its own file and add tests by @lucasnewman in #313
• fix: align disk usage with macOS storage by @Yavaren in #304
• Fix image model empty state and discovery flow by @Yavaren in #300
• Clean up built-in tools listing by @lucasnewman in #314
• Fix model download progress reporting and added progress bar by @Yavaren in #298
• Add conversation forks for prompt editing by @Yavaren in #270
• Feature/shared document extraction by @Yavaren in #309
• Fix stuck download progress indicators by @lucasnewman in #317

New Contributors

• @TheAngryPit made their first contribution in #302

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

• Add Parallel Search to the MCP catalog by @georgeatparallel in #264
• Set up global model folder access by @Lazarus-931 in #266
• Stabilize onboarding model metadata loading by @Lazarus-931 in #274
• Polishing UI by @Lazarus-931 in #277
• removing recurrent option for current chats by @Lazarus-931 in #278
• Stabilize Kits UI layout and status icon by @Yavaren in #280
• Improve model discovery, downloads, and provider branding by @Blaizzy in #262
• UI changes for the MCP page by @Yavaren in #279
• Polish navigation, composers, Artifacts, and model download shutdown by @Blaizzy in #276
• improving markdown rendering in chat by @Lazarus-931 in #285
• Bump version to 0.3.1 by @lucasnewman in #289
• Add model README details panel by @Blaizzy in #290

New Contributors

• @georgeatparallel made their first contribution in #264

Full Changelog: v0.3.0...v0.3.1

v0.3.0

What's Changed

• Make user prompts editable and copyable by @Yavaren in #201
• Draft model browsing, drafter compatibility hints, and a stable decode rate by @Lazarus-931 in #198
• Fix Japanese IME composition being cancelled in the chat input by @Lazarus-931 in #213
• Allow setup to continue while model downloads run in the background by @Lazarus-931 in #132
• fetch model size using async calls, cached by @Lazarus-931 in #216
• Fix dictation shortcut clashing with macOS navigation by @Lazarus-931 in #211
• Fix hub discover pagination: filter at the API + paginate over visibl… by @Lazarus-931 in #217
• Dictation: fall back to on-device system speech recognition by @johntdavies in #190
• Fix various UI issues by @lucasnewman in #220
• Fix user prompt layout breaking when clicked by @Yavaren in #221
• Remove universal label from the header by @lucasnewman in #219
• LLM image generation tool calling feature request #131 by @Yavaren in #215
• Sidebar UI improvements by @lucasnewman in #222
• permission setup during onboarding by @Lazarus-931 in #218
• Mcp hub by @Lazarus-931 in #193
• Semantic search in the artifacts page (embedding endpoint) by @Lazarus-931 in #162
• Artifacts: tag embedding model_type client-side before loading by @Lazarus-931 in #171
• Feature/routines chat first by @Lazarus-931 in #224
• System Monitor: add die temperature, power, and fan sensors by @johntdavies in #187
• Check to make sure all shards are downloaded for sharded VLM/LM models by @lucasnewman in #225
• Skip pre-loading a speech model that cannot load by @Lazarus-931 in #231
• chat title shimmer by @Lazarus-931 in #229
• Report download progress on slow connections by @Lazarus-931 in #236
• Fix stale README details by @Lazarus-931 in #238
• Fix Discover model search responsiveness by @Blaizzy in #241
• Add live download statistics page by @Blaizzy in #240
• Document features missing from the README by @Lazarus-931 in #239
• Fix voice dictation shortcut reliability by @Lazarus-931 in #233
• Restore extension permission controls and hide the Install button by @Lazarus-931 in #234
• Confirm destructive actions by @Lazarus-931 in #248
• Add download growth tracking and history chart by @Blaizzy in #243
• Add meeting transcription suggestions by @Blaizzy in #192
• Fake _receive object was swallowing server requests from external clents. by @dcartman in #237
• Simplify navbar navigation by @Lazarus-931 in #249
• Improve Models search and scrolling responsiveness by @Blaizzy in #242
• Fix debug signing and simplify hands-free shortcut by @Blaizzy in #257
• Feature/hub UI polish by @Lazarus-931 in #247
• Remove Java from custom script tools by @Lazarus-931 in #258
• Feature/browsing platform by @Lazarus-931 in #250
• doc reformat + linking from landing page by @Lazarus-931 in #244
• Improve image model selection for chat tool calls by @Yavaren in #254
• Fold image generation into chat; remove the standalone Images tab by @Lazarus-931 in #227
• Bump version by @lucasnewman in #261

New Contributors

• @Yavaren made their first contribution in #201
• @johntdavies made their first contribution in #190
• @dcartman made their first contribution in #237

Full Changelog: v0.2.2...v0.3.0

v0.2.2

What's Changed

• Disable thinking budget while speculative decoding is active by @Lazarus-931 in #196
• Fix/voice playback 191 by @Lazarus-931 in #197
• Surface server error events from chat completion streams by @Lazarus-931 in #195
• Bump version to pick up new mlx-vlm by @lucasnewman in #202

Full Changelog: v0.2.1...v0.2.2

v0.2.1

What's Changed

• Add provisioning profile by @lucasnewman in #188
• Add release verification script by @lucasnewman in #129
• Keep selection menus open by @Blaizzy in #133
• Make Return activate the default button in confirmation dialogs by @Lazarus-931 in #138
• Generalize the chat tool-call loop into a pluggable tool registry by @C-K-Loan in #141
• Include downloaded image models in the Models installed list by @Lazarus-931 in #144
• Fix settings not persisting (Model Configuration resets, model source folders lost) by @Lazarus-931 in #140
• docs: fix spelling of 'Glamourous' → 'Glamorous' in INTEGRATIONS.md by @lunar-me in #145
• View > Toggle Sidebar was permanently disabled by @C-K-Loan in #142
• Add pinned chats: Pinned and Sessions sidebar sections with a chat co… by @Lazarus-931 in #146
• Fix Japanese and Chinese IME composition in chat input by @Blaizzy in #155
• Fix Discover sort-by-size (invalid Hugging Face API parameter) by @Lazarus-931 in #156
• Add Open Interpreter as an integration by @Lazarus-931 in #152
• Don't bundle training-only Python dependencies by @Lazarus-931 in #153
• Add drag-and-drop pinning, reordering, and bulk actions to the chat s… by @Lazarus-931 in #151
• Add embeddings support by @Lazarus-931 in #147
• Let Return submit destructive confirmation dialogs by @Lazarus-931 in #165
• Show a compact memory-fit warning for oversized models by @Lazarus-931 in #164
• Warn when attaching images to a non-vision model by @Lazarus-931 in #168
• Per-model configuration profiles by @Lazarus-931 in #166
• Add chat session folders to organize the sidebar by @Lazarus-931 in #159
• Apply per-model configuration on the API path by @Lazarus-931 in #167
• Add Artifacts page by @Lazarus-931 in #158
• Add global voice dictation and recording by @Blaizzy in #134
• Chat: drag files from Finder + paste images into the composer by @Lazarus-931 in #174
• Add chat text size control by @Lazarus-931 in #173
• Improve sidebar title / icon alignment by @lucasnewman in #176
• Improve page responsiveness and background work by @Blaizzy in #175
• Sidebar: collapsible sections plus a collapse-all toggle by @Lazarus-931 in #179
• Dictation improvements by @lucasnewman in #177
• Add voice dictation onboarding by @Blaizzy in #180
• Bump version to 0.2.0 (build 3) by @Blaizzy in #181
• Fix release extension bundle identifier by @Blaizzy in #182

New Contributors

• @C-K-Loan made their first contribution in #141

Full Changelog: v0.1.0...v0.2.1

v0.2.0

What's Changed

• Add release verification script by @lucasnewman in #129
• Keep selection menus open by @Blaizzy in #133
• Make Return activate the default button in confirmation dialogs by @Lazarus-931 in #138
• Generalize the chat tool-call loop into a pluggable tool registry by @C-K-Loan in #141
• Include downloaded image models in the Models installed list by @Lazarus-931 in #144
• Fix settings not persisting (Model Configuration resets, model source folders lost) by @Lazarus-931 in #140
• docs: fix spelling of 'Glamourous' → 'Glamorous' in INTEGRATIONS.md by @lunar-me in #145
• View > Toggle Sidebar was permanently disabled by @C-K-Loan in #142
• Add pinned chats: Pinned and Sessions sidebar sections with a chat co… by @Lazarus-931 in #146
• Fix Japanese and Chinese IME composition in chat input by @Blaizzy in #155
• Fix Discover sort-by-size (invalid Hugging Face API parameter) by @Lazarus-931 in #156
• Add Open Interpreter as an integration by @Lazarus-931 in #152
• Don't bundle training-only Python dependencies by @Lazarus-931 in #153
• Add drag-and-drop pinning, reordering, and bulk actions to the chat s… by @Lazarus-931 in #151
• Add embeddings support by @Lazarus-931 in #147
• Let Return submit destructive confirmation dialogs by @Lazarus-931 in #165
• Show a compact memory-fit warning for oversized models by @Lazarus-931 in #164
• Warn when attaching images to a non-vision model by @Lazarus-931 in #168
• Per-model configuration profiles by @Lazarus-931 in #166
• Add chat session folders to organize the sidebar by @Lazarus-931 in #159
• Apply per-model configuration on the API path by @Lazarus-931 in #167
• Add Artifacts page by @Lazarus-931 in #158
• Add global voice dictation and recording by @Blaizzy in #134
• Chat: drag files from Finder + paste images into the composer by @Lazarus-931 in #174
• Add chat text size control by @Lazarus-931 in #173
• Improve sidebar title / icon alignment by @lucasnewman in #176
• Improve page responsiveness and background work by @Blaizzy in #175
• Sidebar: collapsible sections plus a collapse-all toggle by @Lazarus-931 in #179
• Dictation improvements by @lucasnewman in #177
• Add voice dictation onboarding by @Blaizzy in #180
• Bump version to 0.2.0 (build 3) by @Blaizzy in #181
• Fix release extension bundle identifier by @Blaizzy in #182

New Contributors

• @C-K-Loan made their first contribution in #141

Full Changelog: v0.1.0...v0.2.0

v0.1.0

Nativ v0.1.0 is a major step forward for local AI on Apple silicon. This release adds local image generation and editing, expands the integration ecosystem, and makes model management, chat, and server operation faster and more reliable.

Highlights

• Generate and edit images locally — Use the new Images workspace, generate or edit images directly in Chat, paste images from the clipboard, or capture a screenshot. This release also adds a default image model and support for the Mage-Flow family.
• Connect more of your tools — New integrations cover Aider, Goose, Crush, Qwen Code, OpenClaw, Buzz, Zed, Continue, VS Code, Cursor, and JetBrains.
• Manage models with less friction — Nativ now detects models from custom Hugging Face caches and LM Studio, supports custom Hugging Face tokens, keeps downloads running across navigation, allows concurrent downloads, warns when a model may not fit in memory, and surfaces loading progress and failures throughout the app.
• Configure and secure the local server — Choose the server host and port, control the server from the sidebar and Settings, and keep the server API key securely in the macOS Keychain.
• Enjoy a smoother chat experience — Streamed responses are coalesced to keep the UI responsive, scrolling and menus behave more reliably, and conversations can be copied or exported as text.
• See what your Mac and Nativ are doing — Comprehensive macOS system monitoring and richer issue-report diagnostics make performance and troubleshooting easier to understand.

---

What's Changed

• Add warning for models that may not fit in memory by @lucasnewman in #2
• fix: keep model downloads running when leaving the Models page by @Lazarus-931 in #18
• Detect Hugging Face hub cache from HF_HOME/HF_HUB_CACHE by @rcanand in #6
• Add dev build verification workflow for every branch and PR. by @Lazarus-931 in #20
• Detect MLX models installed by LM Studio. by @Lazarus-931 in #21
• Filter logs for metrics requests by @lucasnewman in #36
• Fix software update by @lucasnewman in #37
• feat: allow choosing the local server port by @Lazarus-931 in #19
• Coalesce streamed chat updates to keep the UI responsive by @Lazarus-931 in #48
• Allow custom HF tokens, pull HF_TOKEN in by default by @lucasnewman in #49
• Reserve peak activation in the fit estimate for image models by @Lazarus-931 in #47
• Add Copy Conversation and Export as Text to chat context menu by @Lazarus-931 in #42
• Issue report feature by @Lazarus-931 in #43
• Simplify Developer runtime cards to icon + value by @Lazarus-931 in #41
• Fix Discover pagination showing inconsistent page sizes by @Lazarus-931 in #40
• Add clipboard paste and screenshot capture to chat image input by @Lazarus-931 in #39
• Show recommended models during onboarding by @Blaizzy in #50
• Fix macOS app icon rendering by @Blaizzy in #53
• Fix app icon in updater UI by @Blaizzy in #54
• Show model loading progress across the app by @Blaizzy in #52
• Keep Codex CLI-only and honor dynamic server ports by @Blaizzy in #51
• Add settings page and sidebar server controls by @Blaizzy in #55
• Restore Developer runtime card details by @Blaizzy in #60
• Fix chat scrolling during streaming by @Blaizzy in #61
• docs: note configurable port, HF token/cache and onboarding by @Lazarus-931 in #59
• Fix chat dropdown positioning by @Blaizzy in #63
• Fix header styling to be consistent across all views by @lucasnewman in #68
• Add Images tab for image generation / editing by @lucasnewman in #67
• Add detailed diagnostics to issue reports by @Lazarus-931 in #88
• Add Aider CLI integration by @Lazarus-931 in #72
• Allow attaching the signed appcast to an existing release via workfllow by @Lazarus-931 in #92
• Add Goose CLI integration by @Lazarus-931 in #73
• Add Crush CLI integration by @Lazarus-931 in #74
• Add Qwen Code CLI integration by @Lazarus-931 in #76
• Add OpenClaw CLI integration by @Lazarus-931 in #77
• Add Zed editor integration by @Lazarus-931 in #80
• Add Continue integration by @Lazarus-931 in #82
• Add VS Code integration (guided setup) by @Lazarus-931 in #87
• Add Cursor integration (guided setup) by @Lazarus-931 in #91
• Restore JetBrains integration and complete integration test coverage by @Lazarus-931 in #94
• Invalidate client URLSessions on deinit to fix file-descriptor leak by @Lazarus-931 in #95
• Add support for image generation and editing in chat by @lucasnewman in #93
• Add comprehensive macOS system monitoring by @Blaizzy in #98
• Improve text readability and selection by @Blaizzy in #101
• Refine server menu header status by @Blaizzy in #103
• Add configurable server host by @Blaizzy in #97
• Surface active model downloads outside the Discover list by @Lazarus-931 in #99
• Make the menu bar customizer more discoverable by @Blaizzy in #100
• Replace the floating download indicator with a Models sidebar download arrow by @Lazarus-931 in #105
• Remove the Models sidebar download arrow by @Lazarus-931 in #107
• Show model loading failures to users by @Blaizzy in #108
• Add system, light, and dark appearance selector by @Blaizzy in #112
• Re-add the Models sidebar download arrow by @Lazarus-931 in #110
• Remove the circular spinner from the Models download banner by @Lazarus-931 in #109
• docs: fix space before comma in README.md by @lunar-me in #113
• Chat: centered logo in the empty transcript and an accent hue around the composer in chat page by @Lazarus-931 in #79
• Revert accent color glow around chat composer by @Lazarus-931 in #115
• Fix full-screen sidebar corners by @Blaizzy in #111
• Add a default image model for the Images view by @lucasnewman in #118
• Add Buzz as a guided-setup integration by @Lazarus-931 in #117
• Bump version to 0.1.0 by @Blaizzy in #121
• Add Mage-Flow family model support by @lucasnewman in #119
• Only allow generative (text/vision) models in the chat model picker by @Lazarus-931 in #102
• Store API key in the keychain by @lucasnewman in #122
• Add "Sort by Size" to the model selection menu by @Lazarus-931 in #124
• Add Black Forest Labs branding and improve summary card layouts by @Blaizzy in #125
• Support concurrent model downloads by @Lazarus-931 in #126
• docs: sync README with change by @Lazarus-931 in #83
• Improve sidebar collapse performance by @Blaizzy in #127
• Bump mlx-vlm minimum to 0.6.8 by @Blaizzy in #128

New Contributors

• @lucasnewman made their first contribution in #2
• @rcanand made their first contribution in #6
• @lunar-me made their first contribution in #113

Full Changelog: v0.0.1...v0.1.0