LM Studio Release Notes

1.2.0 - Release Notes

• Faster file suggestions in large projects.
• Faster loading of long side chats.
• Better UI treatment for runtime download progress numbers.
• Consistent sidebar alignment when no sessions exist.
• Session approvals carry over to forks.
• Fixed blurry terminal text at fractional display scales.
• Fixed app updater redownloaded unnecessarily.
• Fixed composer suggestions changing selection before the mouse moves.
• Fixed chat template parsing for Flint models.
• Fixed Mac updates reopening the current version instead of installing the update.
• llama.cpp 2.67.0 extension packs.
• oMLX 0.1.0 engine for Apple Silicon Macs.

Build 2

• Fixed a bug affecting when settings hints are displayed.

Build 1

• Improved LM Link reliability during large transfers.
• Improved runtime discovery reliabilty.
• Added Radeon 8065S detection for ROCm runtime compatibility.
• Fixed chat Markdown formatting bugs.
• Fixed switching to bundled MTP retaining a previously selected draft model.
• [Mac] Added maximum-memory and Neural Engine Prefill controls for Splash models.
• [Mac] Fixed compatibility checks for M5-only Metal runtimes.

1.1.10 - Release Notes

• In-app terminal now available in public Bionic builds.
• oMLX engine for Apple Silicon Macs.
• Excalidraw file viewing previews in chat.
• Faster project loading with side-chat drafts loaded on demand.
• llama.cpp 2.66.0 extension packs.
• SafeTensors model support for speculative decoding.
• Speculative decoding for MLX models.
• Fixed tab-strip scrolling changing direction during trackpad gestures.
• Better runtime compatibility and download selection for M5 Macs.
• Additional compatible LLM backend downloads in Runtime settings.
• Embedding and decision models in the model catalog.
• Experimental MoE expert GPU cache setting for llama.cpp MoE models.
• Restored terminal sessions start automatically when selected.
• Fixed ROCm runtime compatibility detection for AMD Radeon 8065S graphics.
• Code diff syntax highlighting improvements.
• Configurable shell profiles for the in-app terminal.
• More reliable context compaction when conversations exceed model limits.
• Fixed keyboard shortcuts, clipboard handling, saved shell selections, and IME input in the in-app terminal.
• Voice model auto-loading defaults off on Macs with less than 32 GB of memory.
• On-device Read Aloud support on Linux and Windows, in addition to Mac.
• Fixed misleading context limit guidance when multiple requests exhaust model context.
• Fixed multiline shell text being rendered as math.
• Read Aloud button styling now matches other message actions.
• Lower memory usage and detailed progress for large session exports and moves.
• Screen reader support for in-app terminal output.
• Cleaner transcript image previews and message styling.
• Fixed stale draft-model selections when switching to bundled MTP.
• Seven additional offline Read Aloud voices.

1.1.9 - Release Notes

• Lower bandwidth usage for LM Link.
• llama.cpp 2.62.0 extension packs.
• On-device read-aloud for assistant responses.
• Fixed LM Link disconnects during large transfers.
• Session references open in an adjacent split.
• Clearer LM Link errors for incompatible app versions.
• llama.cpp 2.61.0 extension packs.
• Splash runtime compatibility checks for GGUF and MLX models.
• Neural Engine Prefill controls for supported models.
• Phone and tablet icons for LM Link devices.
• Lower context usage from on-demand MCP tool discovery.
• Shared runtime version selections for model formats using the same engine.
• Shared lms CLI installation with LM Studio.
• Higher-resolution Excalidraw canvas exports.
• Maximum memory controls for Splash model loading.
• CSV and TSV file previews.
• Session references by dragging tabs into the composer.
• Faster browsing and search for large session histories.
• Fixed missing error messages throughout Bionic.
• More reliable skill context across retries, copies, and forks.
• Cross-session attachment access for agents.

Build 5

• Fix Windows on ARM CUDA not detected
• Fixed lms compatibility with Bionic.

Build 4

• Support for /v1/systemone Jev (TypeSafe AI) compatible endpoint
• Support for /v1/decisions OpenAI compatible endpoint

Build 3

• Coexistence with Bionic — Run both apps together; Bionic uses LM Studio for local model inference while LM Studio is running. Requires Bionic 1.1.8 and newer.
• Added CUDA 13 support on Windows ARM.
• UI polish

Build 2

• Updated the llama.cpp runtime to 2.50.0 (upstream b11337).

Build 1

• Bug fixes and security hardening.

1.1.8 - Release Notes

• Fix Windows on ARM CUDA not detected
• Fixed lms compatibility with Bionic.
• Draggable Canvas panel
• Support for previewing Word documents, spreadsheets and presentations in-app
• Support for /v1/systemone Jev (TypeSafe AI) compatible endpoint
• Support for /v1/decisions OpenAI compatible endpoint
• Support for previewing HTML files
• Coexistence with LM Studio — Run both apps together; Bionic uses LM Studio for local model inference while LM Studio is running. Requires LM Studio 0.4.26 and newer.
• Decision model load settings, activity status, and cancellation controls.
• Decision model inference and management through the REST API.
• Fixed incorrect MCP connection status reported in chats.
• llama.cpp 2.53.0 extension packs.
• CUDA 13 support on Windows ARM.
• Fixed uneven loader dot spacing at compact sizes.
• Skills in their own composer suggestions category.
• New adaptive Bionic app icon on macOS.
• Fixed onboarding tips reappearing after being dismissed.
• llama.cpp 2.52.0 extension packs.
• Lower disk usage from project history.
• Localization: Bionic native menu items in the selected language.
• More consistent warning, success, and feedback colors.
• Composer suggestions grouped by result type.
• Fixed incorrect macOS versions in session environments.
• Better handling for messages with more than six image attachments.
• Clearer skill references in prompts and previews.
• Fixed side chat sidebar alignment.
• More reliable long sessions with image-aware context compaction that preserves the current environment.
• Localization: approval prompts and previews in the viewer's language.
• llama.cpp 2.51.0 extension packs.
• Fixed unfinished links appearing while responses stream.

1.1.7 - Release Notes

• Improve inline diff UI performance.
• llama.cpp 2.48.0 extension packs.
• Fixed editor toolbar when an edited file is no longer available.
• PDF previews now show comments and tracked changes.
• Fixed gaps around pane content.
• Fixed startup failures caused by corrupted legacy UI state.
• Canvas export to PNG, SVG, and editable Excalidraw files.
• New neutral switch colors and disabled styling.
• Longer image-heavy sessions through automatic context compaction.
• Clickable directory links in Markdown previews.
• More polished PowerPoint layouts, typography, tables, charts, and image placement.
• Fixed Arrow Up navigation in composer suggestions.
• Image menus now include Copy Image and Save Image As.
• Fixed lag when selecting cells in zoomed-out spreadsheet previews.
• Sessions now appear in Cmd + P.
• lms runtime installations now show installation progress.
• Better syntax highlighting in the Markdown editor.
• Cleaner PowerPoint generation with fewer layout and text-overlap issues.
• Markdown previews now match transcript styling.
• Session titles now match the language of your message.
• Nicer skill preview lines for Introspection.

1.1.4 - Release Notes

• Introspection: the agent can now read messages earlier in its own history. Supports @ referencing other sessions.
• Compact file links in the session transcript.
• Better sizing for short labels and values in transcript Markdown tables.
• llama.cpp 2.41.0 extension packs.
• Approval prompts for file operations that exceed checkpoint limits.
• Shorter edited-file summaries for turns with many changes.
• Tool-call argument streaming for compatible local and LM Link models in Chat Completions and Responses APIs.
• Fixed wildcard entries in website access rules.
• Session mentions in the composer, including drag and drop from the sidebar.
• Settings in the macOS app menu, even when no window is open.
• Wider markdown tables in the session transcript.

1.1.3 - Release Notes

• Voice transcription for Linux.
• llama.cpp 2.38.0 extension packs.
• Clearer file and folder drag-and-drop flows in Project Files.
• Agent-created images shown inline in conversations.
• Allow narrower Bionic app window width.
• Consistent dark Bionic app icon on macOS.
• Fixed MLX downloads sometimes stuck at the end.
• Clearer link colors in transcripts and Markdown previews.
• Smoother LM Link connection states and remote model listings.
• Model details when hovering over the collapsed model picker.
• Classic theme when system dark mode is active.
• More reliable non-ASCII output from PowerShell 5.1 commands.
• New Bionic app icon options.
• Fixed Office document tools failing to register.
• Fixed circular symlinks causing file watching loops on Linux.
• German, Spanish, French, Japanese, Brazilian Portuguese, and Traditional Chinese translations.
• Request timeout controls for custom MCP servers.
• Organization-managed MCPs for enterprise deployments.
• MTP speculative decoding for more compatible models.
• Consistent menus across Settings.
• Dark-theme colors for dismiss buttons and skill references.
• Fixed recent project ordering changing from background activity.
• Fixed composer controls shifting when Send changes to Stop.
• Fixed loading models with per-layer GGUF expert counts.
• Fixed guidance dialog layouts with longer translations.
• Fixed tall dialogs dragging the app window.
• Fixed link menus appearing behind other content and cropped suggestion shadows.
• YAML frontmatter shown as file details in Markdown previews.
• Side editor access to files Bionic creates or edits when they need attention.
• Voice transcription on Linux systems with NVIDIA GPUs.

1.1.2 - Release Notes

• Significant UI performance improvements around session loading and tab switching
• Better screen reader support across navigation, dialogs, controls, downloads, and session status.
• Faster session loading, tab switching, and subagent updates.
• llama.cpp 2.33.0 stable extension packs.
• More reliable MLX model detection.
• More reliable browser sign-in.
• File links in Markdown previews.
• More consistent top bar icon buttons.
• Fix Loaded Instances UI showing wrong context length.
• Drag and drop sessions from the sidebar into the main workspace area
• Support for lms load --auto.
• Fixed unread session indicators remaining after opening a session.
• Bionic for Linux (x64 and ARM64).
• Chat API image and text inputs remain in a single user message.
• Chevron controls for empty project folders.
• Side sessions open in the adjacent right pane.