### What's New • Added Attach Files, Attach Notes, Attach Knowledge, and Skills sections to the + menu in chat — quickly attach server-uploaded files, notes, or knowledge bases, and toggle skills directly from the compose menu. ### Improvements • Reasoning effort parameter now includes "xhigh" as a preset option and a "Custom" free-text mode for any value a provider supports. • Many qol updates. ### Bug Fixes • Fixed a frosty/translucent bar appearing at the top of the chat screen when scrolling — the navigation bar background now stays fully transparent at all scroll positions.

### Improvements • Minor visual improvements

### What's New • Added "Using Entire Document" toggle in the file attachment preview — tap a file pill to choose between Focused Retrieval (default) or sending the full document content inline. ### Bug Fixes • Fixed "Save to Photos" silently doing nothing — the app now requests Photos permission on first save and opens Settings if access was previously denied.

### What's New • Added My Defaults settings to customize account level system prompt and advanced parameters. ### Improvements • Improved the animation on app launch ### Bug Fixes

### What's New • Read Aloud (Server TTS only) now continues playing your response when you put the app in background allowing you to listen to long responses while navigating other apps on your phone. ### Improvements • Improved scroll FABs: The ↑ button jumps to the previous question • App and server update checks now happen when returning from background as well. • Terminal button in chat input now only appears when the selected model has terminal capability enabled — configure per model in the workspace model editor. Matches openwebui changes. ### Bug Fixes • Fixed currency amounts being read incorrectly by TTS • Fixed text selection popup (Ask/Explain menu) auto-closing after 2 seconds • Fixed regenerated responses not fully settling markdown

### What's New • Added "Ask" and "Explain" buttons to the text selection menu in assistant messages — select any text to quote it into the input box or instantly ask for an explanation. • Added delete chat button to the chat toolbar — tap ··· in any chat and choose Delete to permanently remove it (with confirmation). • Added Markdown Rendering toggles in Settings → Chat Behavior — independently control whether markdown is rendered in your messages and AI messages. • Added scroll to top navigation fab ### Improvements • Tapping the new chat button now transitions smoothly instead of abruptly flashing the UI. • Sources bar now shows a +N badge when a response has more than 3 citation sources, keeping the bar compact while making all sources accessible. ### Bug Fixes • Fixed app becoming unresponsive to taps when returning from background with the sidebar partially open mid-swipe

### Bug Fixes • Fixed terminal shell failing to reconnect automatically after being idle — it now correctly detects a dead session and establishes a fresh connection instead of silently giving up. • Fixed app becoming unresponsive (buttons not tappable) after returning from a long background session

### What's New • Added "Manage Ollama" sheet — pull models from Ollama.com with real-time whole-model progress, update all installed models at once, delete models, create custom models from a modelfile, and upload local GGUF files directly to your Ollama server, all from within the app.

### What's New ### Improvements • Smoother transitions throughout the app. • Fixed iPad sidebar chat list bouncing/hopping when tapping a chat to open it. ### Bug Fixes • Fixed terminal WebSocket creating multiple orphaned sessions when reconnecting after switching apps or reopening the terminal panel.

### What's New - Added models tab under Admin Console settings - You can now manage your models from the app directly. ### Bug Fixes - Fixed app freezing when switching to a chat containing images. - Fixed intermittent crash when opening model selector, model editor, tool valves, or chat parameters sheets.

### What's New • Complete redesign of open terminal shell architecture — now powered by SwiftTerm, native terminal experience using websockets instead of http polling. It should feel snappier and realtime. ### Bug Fixes • Fixed typing indicator appearing as a large full-width element while the AI is generating a response.

### What's New • Added Share Extension — share URLs, files, images, and documents from any app directly into Open Relay. • Added Face ID / Touch ID sign-in — after signing in once, tap the Face ID button on the login screen to authenticate instantly. Also allows full passwords support now! • Native Text Selection on macos - When using ipad version on mac, you are now freely able to select text just like you would in any other app. ### Improvements • Further stability improvements when loading into chats. • Improved password manager and iCloud Keychain AutoFill support on the sign-in screen. ### Bug Fixes • Fixed the app signing you out when your server is temporarily unreachable — transient connectivity issues should no longer clear your session.

Bug Fixes • Attempted fix for "Sign In Failed" error when signing in with Microsoft Entra ID • Fixed multiple crashes affecting streaming responses.

### Bug Fixes • Fixed sidebar sections (Pinned, Today, Yesterday, etc.) resetting to expanded on every app launch — collapse state now persists across launches. • Fixed various server-level enabled/disabled settings.

### Improvements • Images embedded directly in responses now display inline with no raw data visible during generation ### Bug Fixes • Fixed the chat scrolling back to the bottom after streaming ends if the user had manually scrolled up during the response.

### What's New • Added a "save as permanent" button to temporary chats • Images embedded directly in responses now display inline — tap to view fullscreen with pinch-to-zoom, or hold to save to Photos or share ### Improvements • Significantly reduced false "server disconnected" alerts • Returning to the app from the background should reconnect instantly • Added a subtle haptic when a response finishes streaming ### Bug Fixes • Fixed a race condition where streaming auto-scroll would randomly disengage mid-response.

### What's New • Added Message Queue — when enabled in Chat Settings, messages sent while the AI is responding are queued and sent automatically when the response finishes. Each queued message can be sent immediately, edited, or removed from the queue. ### Improvements • Added a share button to inline Rich UI embeds. ### Bug Fixes • Fixed double requests (false emoji generation) being sent on every request.

### What's New • Hold on any message text to select and copy — long-press text selection now works alongside the existing double-tap. ### Improvements • Input bar buttons and icons now scale with your Accessibility UI Scale setting in both chat and channel views. • Code block/native visualization performance has been drastically improved - While streaming code blocks that could contain 600+ lines, UI remains responsive and lag free! • Markdown underlying improvements. ### Bug Fixes • Fixed streaming haptics not firing during streaming. • Fixed the message composer requiring multiple taps to open the keyboard — tapping anywhere on the input bar now reliably focuses it. • Fixed a jitter when dismissing the keyboard by scrolling — the input bar now glides down in perfect sync with the keyboard instead of snapping after it closes. • Fixed continuous main-thread animation frames competing with the WebContent render during HTML streaming, causing visible lag when JS-heavy content streams in.

### Bug Fixes • Fixed app not showing the update notification for new versions • Fixed pending accounts incorrectly entering the app on relaunch • Fixed pipe/function models throwing error due to latest owui changes

### Bug Fixes • Fixed update notification popup appearing every time the app launches — once dismissed, it won't reappear for that version; the update icon in the sidebar stays visible so you can tap it anytime. • Fixed widgets not rendering properly.

### What's New • Terminal now has a fullscreen mode — tap the expand button in the terminal toolbar to open a full-screen terminal view with more room to work. • Terminal now shows a quick-action shortcut bar with Tab, history up/down (↑ ↓), and Clear buttons for faster keyboard-free input. ### Improvements • Terminal shell now starts warming up as soon as you enable terminal access, so input is ready instantly when you open the panel. ### Bug Fixes • Fixed terminal output displaying raw ANSI escape codes (e.g. [1m[96m…[0m) instead of clean text. • Fixed terminal access (terminal_id) still being included in chat requests even after disabling the terminal toggle. • Fixed server update changelog showing raw HTML code blocks or being invisible — now renders as native SwiftUI views with bold titles and color-coded category labels. • Server update sheet now shows the server's favicon as the icon (with a server.rack fallback if unavailable). • Fixed conversation tags not appearing — tags were being read from the wrong field in the server response; now correctly reads from the updated server format. • Fixed conversation timestamps (created/updated dates) showing the wrong time after loading a chat — the server sends them as integers which weren't being handled correctly.

What's New - Long-press any chat inside a folder to enter selection mode — multi-select, then remove from folder, move to another folder, or delete in bulk. - Added "Move to Folder" button in the main chat list's selection toolbar — select multiple chats and move them all into a folder at once. - Added server update notifications — the app now checks your Open WebUI server for available updates on launch. A "Check for Server Updates" button is also available in About → Server. - Image size reduction - 5mb cap on vision models cause error if the image is too big. - Native visualizations now support audio output. Improvements - Minor code block performance improvements. - Removed "Clear Local Cache" option from Privacy & Security settings as Storage Menu covers it all. Bug Fixes - Terminal now runs as a persistent bash session with proper stdin support; suppressed harmless startup warnings - Fixed terminal remaining active for the model after disabling it — the model no longer receives terminal access when the toggle is off - Fixed admin-disabled terminal servers appearing as available and being silently used in requests - Fixed multi-server terminal: switching servers with the toggle off no longer leaves a stale server selected that could re-activate on the next enable - Fixed folders only showing up to 10 chats — now loads all chats across all pages smoothly - Fixed typing indicator animation - Fixed update notice incorrectly triggering based on GitHub tags instead of the actual App Store version.

What's New - Added dedicated Input Box Text size slider in Accessibility settings - independently scale the font in the chat and channel message composer. - A complete re-write of the streaming pipeline - Massively improved streaming performance. App should feel fully responsive while streaming. Improvements - Switching between chats should feel smoother. - Screen now stays on while TTS is reading a response aloud, just like during voice calls. - Voice calls now respect the admin-configured Voice Mode prompt from OpenWebUI's interface settings - the model automatically receives the right system prompt to keep responses concise and speech-friendly. Bug Fixes - Fixed nested code blocks rendering as plain text. - Fixed citation source icons showing as letter avatars instead of favicons. - Known Issue: Inline visualizer plugin is not correctly working with the new re-write and needs a bit more work. Continue using the native visualizer which works perfectly and much better than pluggin. In future, the pluggin support may be dropped as it requires lot of processing since its written for the webui (includes iframe code in the tool block) and causes lag and has no real benifit over the native visualization.

What's New - Added user valve editor for tools — tap the gear icon next to any tool in the tools picker to configure its settings. - Added in-app update notice — the app checks for new versions on every launch and shows a sheet with release notes; tapping "Later" dismisses it and leaves an update icon next to the New Chat button in sidebar so you can reopen it anytime. - Added star button to every tool in the tools picker — tap the star to instantly pin it as a quick-action pill in the chat input, no settings detour required. Improvements - Tool call results now display with full syntax highlighting and virtual windowing — only the visible portion is rendered, matching how regular code blocks work. - Streaming responses now scroll more smoothly with less jank, reduced CPU usage, and no animation stutter. - Dramatically improved performance for reasoning/thinking models — the app no longer gets slower as the model thinks longer. - Channel deletes, edits, reactions, and pins from other devices now appear live without needing to refresh. - Deleting a channel message now animates out instantly instead of waiting for the server. Bug Fixes - Fixed TTS audio not playing through earbuds or wired headphones in both chat read-aloud and voice calls. - Fixed regenerated responses not showing server errors in the chat bubble. - Fixed the chat not scrolling to the regenerating message — regenerating a response now animates the view to the new message the same way sending does. - Fixed non-toggleable filters incorrectly appearing under Default Filters in the model editor. - Fixed toggle-filter functions in the tools picker always turning on when globally enabled, ignoring the model's configured default state. - Fixed crash in channels when deleting a message. - Fixed channel context menu appearing off-screen when the keyboard was open. - Fixed new chats always starting with the last-used model instead of the server-configured default model. - Fixed Settings Default Model changes not applying to new chats opened in the same session. - Hidden models (disabled by an admin in OpenWebUI) no longer appear in the model picker or the Default Model setting. - Fixed thinking blocks breaking the "Explored N" tool-call grouping on reasoning models — all tool calls now collapse into a single pill as expected.

Drastically improved app’s performance while streaming!