Built for iPhone Duo — Unfolded, the editor and live preview share the screen and split around the fold. In tabletop posture, the preview stands on the upper half while you type on the lower half. Updated diagram engines — Mermaid 12, PlantUML, and Graphviz are updated. Your existing Mermaid diagrams keep their familiar look; PlantUML sequence diagrams use PlantUML's newer default layout. Smaller download — The PlantUML runtime is less than half its previous size. Fewer render errors — Diagrams that mention a URL in a label or link now render instead of being rejected, including ones from Generate and Fix. Links stay inside the preview. Diagram Cloud — Updated sign-in components for more dependable session restore.

Ten languages — The interface is now translated into English, Spanish, French, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, Traditional Chinese, and Italian. Translated template library — Every template category, name, and description is translated, so browsing for the right starting point works in your language too. Translated Pro workspace — Generate, Ask, and Fix, the local LLM setup, and Diagram Cloud sign-in and sync all read in your language, including their status and error messages. Per-app language picker — Settings ▸ Appearance ▸ Language lets you run PrivyMermaid Pro in a language of its own, independently of the rest of your device. Choose System to follow your device setting. Sensible language matching — Regional variants land where you would expect: Traditional Chinese for Taiwan, Hong Kong, and Macau, and Brazilian Portuguese for other Portuguese regions. Your diagrams stay yours — Diagram source is never translated, and existing diagrams, projects, and cloud templates keep working unchanged when you switch languages.

Ten languages — The interface is now translated into English, Spanish, French, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, Traditional Chinese, and Italian. Translated template library — Every template category, name, and description is translated, so browsing for the right starting point works in your language too. Translated Pro workspace — Generate, Ask, and Fix, the local LLM setup, and Diagram Cloud sign-in and sync all read in your language, including their status and error messages. Per-app language picker — Settings ▸ Appearance ▸ Language lets you run PrivyMermaid Pro in a language of its own, independently of the rest of your device. Choose System to follow your device setting. Sensible language matching — Regional variants land where you would expect: Traditional Chinese for Taiwan, Hong Kong, and Macau, and Brazilian Portuguese for other Portuguese regions. Your diagrams stay yours — Diagram source is never translated, and existing diagrams, projects, and cloud templates keep working unchanged when you switch languages.

Ten languages — The interface is now translated into English, Spanish, French, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, Traditional Chinese, and Italian. Translated template library — Every template category, name, and description is translated, so browsing for the right starting point works in your language too. Translated Pro workspace — Generate, Ask, and Fix, the local LLM setup, and Diagram Cloud sign-in and sync all read in your language, including their status and error messages. Per-app language picker — Settings ▸ Appearance ▸ Language lets you run PrivyMermaid Pro in a language of its own, independently of the rest of your device. Choose System to follow your device setting. Sensible language matching — Regional variants land where you would expect: Traditional Chinese for Taiwan, Hong Kong, and Macau, and Brazilian Portuguese for other Portuguese regions. Your diagrams stay yours — Diagram source is never translated, and existing diagrams, projects, and cloud templates keep working unchanged when you switch languages.

Ten languages — The interface is now translated into English, Spanish, French, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, Traditional Chinese, and Italian. Translated template library — Every template category, name, and description is translated, so browsing for the right starting point works in your language too. Translated Pro workspace — Generate, Ask, and Fix, the local LLM setup, and Diagram Cloud sign-in and sync all read in your language, including their status and error messages. Per-app language picker — Settings ▸ Appearance ▸ Language lets you run PrivyMermaid Pro in a language of its own, independently of the rest of your device. Choose System to follow your device setting. Sensible language matching — Regional variants land where you would expect: Traditional Chinese for Taiwan, Hong Kong, and Macau, and Brazilian Portuguese for other Portuguese regions. Your diagrams stay yours — Diagram source is never translated, and existing diagrams, projects, and cloud templates keep working unchanged when you switch languages.

PlantUML support — Create, import, edit, preview, and export PlantUML diagrams alongside Mermaid and D2. Graphviz support — Create, import, edit, preview, and export Graphviz DOT diagrams alongside the other supported languages. Four diagram languages in one workspace — Start a new diagram in Mermaid, D2, PlantUML, or Graphviz. AI assistance for more diagram types — Use Ask, Fix, and Refactor with PlantUML and Graphviz source, as well as Mermaid and D2. Expanded file imports — Open PlantUML files such as .puml, .plantuml, .pu, and .uml, plus Graphviz files such as .dot, .gv, and .graphviz. Format-aware sharing and exports — Export PlantUML and Graphviz diagrams as SVG, PNG, PDF, or GIF, or share their source files in the matching format. Improved source highlighting — PlantUML and Graphviz source now have language-aware syntax highlighting.

D2 diagram support — Create, import, edit, preview, and export D2 files alongside Mermaid diagrams. Projects and files — A new native Projects tab lets you create folders, search files, rename or delete diagrams, and move them between projects or Ungrouped. Diagram Cloud sign-in — Use a secure email link, GitHub, or Google to access Diagram Cloud. No keys or provider configuration are exposed in the app. Automatic cloud sync — Once signed in, local saves automatically update the corresponding Diagram Cloud diagram in your default Ungrouped project. Local work remains safe and available offline. Cloud-aware templates — Browse templates with purpose tags and a clear Mermaid or D2 language label.

D2 diagram support — Create, import, edit, preview, and export D2 files alongside Mermaid diagrams. Projects and files — A new native Projects tab lets you create folders, search files, rename or delete diagrams, and move them between projects or Ungrouped. Diagram Cloud sign-in — Use a secure email link, GitHub, or Google to access Diagram Cloud. No keys or provider configuration are exposed in the app. Automatic cloud sync — Once signed in, local saves automatically update the corresponding Diagram Cloud diagram in your default Ungrouped project. Local work remains safe and available offline. Cloud-aware templates — Browse templates with purpose tags and a clear Mermaid or D2 language label.

Custom AI endpoints — Connect to Ollama, LM Studio, or any custom OpenAI-compatible server by URL, with an optional API token stored securely in the device Keychain. Ask with images — Attach screenshots or photos to your questions and chat with vision-capable models about your diagrams. Formatted answers — AI responses now render as Markdown — headings, lists, code blocks, and tables — for easier reading. Reasoning-model support — Output from thinking models shows a collapsible "Thinking" section, keeping the final answer front and center. Reliable local-network connections — Reach your LAN or localhost Ollama / LM Studio server over HTTP without connection errors. Cleaner Settings & About — Simplified privacy summary, a built-in Privacy Policy link, and a dedicated About screen. Still 100% private — Your diagrams and prompts go only to the AI endpoint you configure. No cloud, no telemetry.

Custom AI endpoints — Connect to Ollama, LM Studio, or any custom OpenAI-compatible server by URL, with an optional API token stored securely in the device Keychain. Ask with images — Attach screenshots or photos to your questions and chat with vision-capable models about your diagrams. Formatted answers — AI responses now render as Markdown — headings, lists, code blocks, and tables — for easier reading. Reasoning-model support — Output from thinking models shows a collapsible "Thinking" section, keeping the final answer front and center. Reliable local-network connections — Reach your LAN or localhost Ollama / LM Studio server over HTTP without connection errors. Cleaner Settings & About — Simplified privacy summary, a built-in Privacy Policy link, and a dedicated About screen. Still 100% private — Your diagrams and prompts go only to the AI endpoint you configure. No cloud, no telemetry.

Custom AI endpoints — Connect to Ollama, LM Studio, or any custom OpenAI-compatible server by URL, with an optional API token stored securely in the device Keychain. Ask with images — Attach screenshots or photos to your questions and chat with vision-capable models about your diagrams. Formatted answers — AI responses now render as Markdown — headings, lists, code blocks, and tables — for easier reading. Reasoning-model support — Output from thinking models shows a collapsible "Thinking" section, keeping the final answer front and center. Reliable local-network connections — Reach your LAN or localhost Ollama / LM Studio server over HTTP without connection errors. Cleaner Settings & About — Simplified privacy summary, a built-in Privacy Policy link, and a dedicated About screen. Still 100% private — Your diagrams and prompts go only to the AI endpoint you configure. No cloud, no telemetry.

* Added Mermaid syntax highlighting in the Pro editor, covering diagram types, directions, sequence keywords, styling directives, Gantt/data keywords, Git graph, requirement diagrams, C4, architecture, block syntax, and comments. * Fixed an issue where the current Mermaid file might not render automatically when opening the app for the first time. * Fixed PNG export naming so exported .png files now use the current Mermaid document name. * Improved editor and preview reliability during app startup, document switching, and export workflows.

Full Template Library — Explore built-in Mermaid templates across Software Development, E-commerce, Payment, Search, Messaging, Content, Live Streaming, and more. More export formats — Export diagrams as Mermaid source (.mmd), SVG, PNG, PDF, or animated GIF from the preview toolbar. Animated GIF export — Mermaid edge animations can now be captured and shared as GIFs, making dynamic flowcharts easier to demo. Improved export quality — PNG and PDF exports now use more reliable sizing and background handling for cleaner results. Rendering improvements — Preview rendering is more robust during quick updates, with better support for animated edges. Private AI stays local — Continue generating, asking, and fixing Mermaid diagrams with your local AI setup in PrivyMermaid Pro. Faster starts for complex diagrams — Pick a template, customize it, and then refine it further with Pro’s local AI workflow.

More export formats — Export diagrams as Mermaid source (.mmd), SVG, PNG, PDF, or animated GIF from the preview toolbar. Animated GIF export — Mermaid edge animations can now be captured and shared as GIFs, making dynamic flowcharts easier to demo. Improved export quality — PNG and PDF exports now use more reliable sizing and background handling for cleaner results. Rendering improvements — Preview rendering is more robust during quick updates, with better support for animated edges. Private AI stays local — Continue generating, asking, and fixing Mermaid diagrams with your local AI setup in PrivyMermaid Pro.

- **LM Studio support** — Choose Ollama or LM Studio as your local LLM engine. Set a custom endpoint for each and pick your model from Settings. - **First-launch LLM setup** — New Pro users see a short onboarding screen: configure your LLM (Configure) or skip and set it up later (Skip). If you skip, the Pro panel shows a clear “not configured” message with a link to Settings. - **Streamlined Settings** — LLM (Pro) section with Engine picker (Ollama / LM Studio), endpoint fields, model list, and “Refresh models.” Prompts are no longer in Settings. Open Source Licenses and About are in Settings under Legal and at the bottom. - **Stability** — License viewer resource path fixed so the in-app Open Source Licenses page loads correctly.