Changelog: March 13–30, 2026
Three weeks, one massive update.
Upgraded Super Mind mode is live – all five AIs now fuse their thinking into one ultimate answer. Smart Visualizations let you generate and download charts directly from conversations. AIs now remember previous turns natively, onboarding learns what you need before you even ask, and you can personalize how every AI talks to you. Plus: push notifications, BYOK support, a full mobile redesign, and dozens of fixes across the board.
Major New Features
- Super Mind Mode (Fusion) — Upgraded orchestration mode that runs all 5 AIs in parallel and provides you with the answer based on all five AI replies. The ultimate answer.
- Automatic Smart Visualizations — AI responses now automatically include relevant charts and graphs — bar charts, line charts, heatmaps, and tables — whenever the data calls for it. Multiple charts per message, downloadable as PNG with transparent backgrounds, and automatically embedded in your Master Document exports. A dedicated “Visuals” tab in the sidebar gives you a gallery view of everything generated.
- User Requested Smart Visualizations — You can now, directly from the thread, request the creation of graphs and charts based on the data in the AI messages, for instant PNG download. You can embed it in your documents, reports, and other things. No more struggling with Excel. Just grab the paragraph with data that you like, or copy-paste your own raw data, and Suprmind will in two or three seconds, generate the selected graph type in the selected color pattern and give you the option to download it as PDF, PNG, or SVG.
- Enhanced Conversation Continuity — OpenAI, Grok, and Gemini, in addition to our Context Fabric, also maintain server-side conversation memory via chaining/Interactions APIs. This results in more natural conversation flow and even better context persistence for longer threads.
- User Personalization System — New Settings tab where you can describe your role/biography/preferences, so AIs know with whom they are talking to, and so they can use your projects or information as examples or solutions, and generally improve the quality of communication.
- Bring Your Own Key (BYOK) — To further increase your usage limits, you can use your own API keys for any provider. Your usage is tracked separately and doesn’t count against your plan limits.
- “All Responses Completed” Push Notifications — Response-ready alert for when all five are finished responding, so that you in the meantime can do work in other tabs without the need to monitor the conversation. Privacy policy updated.
- Streaming Adjudicator — The Adjudicator decision brief now appears section by section as it’s written, so you can start reading immediately instead of waiting for the full analysis to complete.
- Mobile UI Overhaul — Preset prompts are now swipeable pills at the top of the screen. Cleaner toolbar, wider sidebar that extends to the screen edge, and compact mode pills that fit in a single row. Overall a much tidier experience on phones and tablets.
- Streaming Speed Control — You can now control how fast AI responses and Master Documents render on screen — useful if you prefer reading at your own pace or want to skip ahead faster.
- Better Master Document Exports — Improved formatting quality across PDF and Word exports — cleaner headings, properly aligned blockquotes, correct table widths, and fixed character rendering for non-Latin languages.
- Jump to Latest Line — A floating button appears when you scroll up in a long conversation, letting you jump back to the newest message in one click.
Improvements
- Claude Prompt Caching — Claude now reuses previously processed context across sequential, debate, and fusion modes, resulting in faster responses and lower costs on longer conversations.
- Smarter AI Prompts — AIs now respond in your language automatically, reference themselves more naturally across turns, and produce fewer hallucinations in Scribe notes. Overall response quality is noticeably improved.
- Custom Provider Order — Choose which AI responds first in Sequential mode from Settings → Modes. Technical model IDs are hidden — you just see the AI names.
- Faster First Response — The first AI reply in a new conversation now arrives noticeably faster thanks to optimized startup processing.
- Higher Output Limits — All AIs can now produce significantly longer responses, supporting more detailed and comprehensive answers for complex questions.
- Settings Redesign — Cleaner layout with labels inside inputs, side-by-side plan comparison cards in billing, and a redesigned desktop settings dropdown.
- Faster Master Documents — Master Documents now generate faster and auto-scroll as content appears, so you can start reading while the document is still being written.
- Subscription Management — Replaced broken cancellation popup with native flow, and on the plan page, you can directly from the app update your payment details.
- Intercom → Sidebar — Moved from floating bubble to sidebar item, to stop it from covering parts of the screen, especially on mobile devices. It’s still fully active and available for support purposes.
Did you know?
You can queue follow-up messages while AIs are still responding – just type and hit Enter. Your messages will be sent automatically once the current turn finishes.
Combine that with push notifications, and you can warm up the AI team in the background while you do other work. When you come back, they’re primed and ready.
–
