Release Notes - Version 2.6.0

📊 AI Usage, Rebuilt (Improvement)

The AI Usage page is about your plan: A plan card shows its name and how many clouds and people share it, with "You, today" and "This plan, this week" meters and their reset times
Know what happens at a limit: A new panel explains, for your plan, when chat carries on with Emerge Fast & Free, what pauses, and how Extra AI Tokens keep premium models running
Plans without limits: Plans with no daily or weekly limit get a clear card that says so
Who's using the most: Plan Admins and the Account Owner see the plan's weekly usage split by its top four people and everyone else

👥 Each Person's Usage in Plan and Billing (New Feature)

Usage beside every member: The Members tab shows each person's share of the weekly limit, how much of their own daily limit they've used today, and when they were last active
Weekly and daily views: Switch views, filter by High, Medium, or Low usage or by who's reached their daily limit, then search and sort
Built for big teams: Search, filters, and paging appear as your plan grows
Percentages only: Nobody's chats are ever shown, and only the plan's Account Owner sees these columns

📇 Records Beside Your Chat (Improvement)

Side by side: A record or task you open from a chat now sits next to the conversation, so you can read both at once
Tasks open in full: A task opened from chat shows its complete detail view, comments included
The AI can read what you have open: Ask about the record beside the chat and Emerge.ai reads it directly
Update a record from chat: Ask for a change and Emerge.ai proposes it on a card showing each field's current and new values. Nothing is saved until you approve, and if the record changed since it was read, nothing is overwritten
Know where it's looking: The status line names the app being searched, and a new record's card names where it will be created

🎨 Document Themes, Headers, and Footers (New Feature)

Six ready-made themes: Default, Emerge, Corporate, Forest, Slate, and Sunset each set heading colours, table header fills, and the font
Headers and footers: Add a header or footer from the insert menu. They stay out of the way until you type in them
Exports keep the look: Themes carry into HTML, PDF, image, and Word exports, and headers and footers into HTML, PDF, and image exports
Ask the AI for a theme: Emerge.ai can apply a theme while it writes or revises a document
Images straight into documents: Ask Emerge.ai to generate an image and put it in a document in a single message, or to use an image already in the chat

📄 Document Improvements

See where a document lives: The editor header shows the document's Drive folder, or AI Files for documents Emerge.ai made. Click it to move the document anywhere in your Drive
Word exports: Keep your page colour, images, and readable text on dark pages, and start on the first page
HTML exports keep their images: Images are embedded in the file, with a note if any couldn't be
PDF exports: Inline code, headings, and strikethrough keep their style, and page breaks fall between blocks
Download rich-text documents as Word: Downloading several files from Drive turns rich-text documents into Word files, and Export lets you pick the format for a single one
AI compose pane polish: The pane beside your document shares the editor's width, has its own close button, and follows the AI's edits on a brand-new document

✅ Tasks Improvements

Attach files to a task: Upload files or pick them from Drive, or drop them anywhere on the task. They appear as cards you can preview or remove, and you can add them while creating the task
Actions within reach: Complete and Add Comment sit in a bar at the bottom of an open task, next to its due date
Sort by tag or owner: Group your list by tag or owner, with the soonest due first in each group
Helpful empty screens: An empty list tells you whether you have no tasks yet or none in the current view
Due dates from chat: Asking `@@Tasks` for something "due Friday" sets the task's due date

🏷️ Cloud Tags (New Feature)

One place for your cloud's tags: Cloud Settings → Tags lets admins create, rename, recolour, describe, and delete tags. Everyone else can see the list
Colourful, easier picking: The tag picker shows tags as coloured pills, lets you choose a colour as you create one, and opens upward when there's no room below
Clean deletes: Deleting a tag removes it from every record and task in the cloud
Manage from anywhere: A "Manage tags" link in the picker opens the settings page

🧠 Skills Improvements

Skills read their whole folder: Emerge.ai opens a skill's reference files and templates when it needs them, so large skills leave more room for the conversation
See what it's reading: The status line names the skill and file, like "reading Brand Voice › templates/email.md"
The right controls in the right file: Skill and model pickers appear when you edit a skill's main file, where they take effect
Back to your chat: Opening skill settings from a skill in a chat gives you a Back button to return

🔌 JobDiva Connector (New Feature)

Set up once for everyone: A Cloud Owner or Admin registers your organisation's JobDiva account, and everyone in the cloud can use it with no separate connection step
Search candidates and job orders: Mention `@@JobDiva` in a chat to look them up
Creating asks first: Adding a candidate or job order waits for your approval
Checked before saving: Your JobDiva details are tested when you save them, with a clear message if the Client ID isn't a number

🤖 New AI Models

Claude Opus 5.5: Anthropic's fastest, most capable Opus yet, for agentic coding and long-horizon reasoning. It takes in around 750,000 words at once
Claude Sonnet 5.5: The best balance of speed and intelligence, taking in around 750,000 words at once. Built-in assistants now run on it
Claude Haiku 5.5: The fastest Claude, for high-volume tasks like sorting, extracting, and routing, taking in around 750,000 words at once
GPT-6 Sol: OpenAI's frontier GPT-6 model for complex coding and agentic work, taking in around 790,000 words at once
GPT-6 Luna: A fast, low-cost GPT-6 model for focused, high-volume tasks, with the same reach as Sol
Grok 4.7: xAI's latest frontier agentic model for coding, research, and long-running tasks, taking in around 375,000 words at once

🔐 Account Security

Sign out everywhere else: Changing your password offers "Sign out of all other devices", on by default
Confirm it's you: Deleting your account or a cloud asks you to sign in again if it's been a while
One account, two ways in: Signing up with an email that already has an account, such as one created with Google, links your new password to it, and you can sign in either way

💬 Chat and App Improvements

Upload voice recordings: Attach recordings you already have, like an iPhone Voice Memo, in chat or Drive. M4A, MP3, WAV, AAC, and WebM files are supported
Click-to-edit, everywhere: Chat titles, file names, contact fields, task titles, and table cells all edit the same way: Enter saves and Escape cancels
A new sidebar shortcut: Show or hide the sidebar with ⌘ \ on Mac or Ctrl \ on Windows, so ⌘ B stays bold in the message box. Hover the sidebar button to see it
Personal and Team tabs: Both tabs now appear in the sidebar on every plan, including one-person plans
Tables fill the page: Tables in AI answers and pasted tables stay full width when you edit them
Search boxes: Clear any search with Escape or its clear button

💳 Plan and Billing Improvements

Seats and clouds at a glance: One bar each shows how many seats and clouds are used, open, or extra, with buttons to add seats, add cloud slots, or remove extras
A tidier role picker: Member roles use a compact badge-style menu in Plan and Billing, Manage Members, and the invite screens
One clear header: The Plans page opens under a single "Change your plan" title

👋 A Guided Start for New Accounts (New Feature)

Set up in under a minute: New accounts confirm their name, add a photo, and name their Cloud, with a live preview of the app filling in as they type. It picks up on any device, web or desktop
A welcome offer for new accounts: Setup may end with a limited-time discount on a paid plan, with a countdown in the app header. Claim it before it runs out
Promotion links: Open one to claim a free month of a paid plan with no card needed, or to save a discount for when you choose a plan

🖥️ Emerge.ai Desktop (New Feature, Beta)

Emerge.ai on your computer: Download the desktop app for Mac (Apple Silicon) or Windows (64-bit) from emerge.ai/desktop. It has the same chats, projects, Drive, Contacts, and Tasks as the web app
Local models: Run AI right on your computer — see below!
Sign in and upgrade through your browser: Google, Microsoft, and other sign-ins and Stripe checkout open in your browser, then bring you straight back to the app
Updates itself: New versions download in the background and install when you restart or step away, reopening on the page you were on. You can also check for updates from the menu or Settings
Downloads with progress: Files save with a progress note you can cancel and a Show in Finder (or Show in folder) button when done
Local clouds are coming soon: Keep projects, chats, and files in a cloud that lives on your own machine

💻 Local Models on Desktop (New Feature, Beta)

Run AI on your own computer: On paid plans, the desktop app can chat with open models running on your computer through the free Ollama app, or a model server you already use like LM Studio, with no cost per message
Recommended for your computer: Emerge.ai checks your computer's memory and suggests the best models for it, with download sizes and speed estimates, and shows the license and terms before anything downloads
Pick Cloud or Local per chat: Switch between cloud and local models in the model picker, choose how much the model thinks first, and attach images to models that can see them. Skills work with local models, and apps and connectors are coming soon
Doesn't count toward your limits: Local model replies are free and don't use your plan's AI allowance. Your chat history still saves to your cloud like any other chat
Admins stay in control: Cloud Settings has a switch to allow or turn off local models for a cloud

🐛 Bug Fixes

Getting started: Accounts whose sign-up didn't finish can create their personal Cloud, and a member whose invite didn't fully complete gets the cloud's plan
Chats and projects: Messages can be sent with only an attachment, Linked Chats lists the right chats, restored chats reappear in the sidebar, and apps picked on the New Chat screen stay off your first message
Sharing follows the chat: Images and documents in a chat open for people added to its project and close for people removed, and moving a chat or project to another cloud keeps every document version and image
Files and Drive: AI documents in a project's Files tab open when clicked, AI files saved to a folder appear there, favourite folders open from Favorites, and skill folders can share a name across skills
Records and tasks: Rich-text fields load their lists and inline code, a new task starts with an empty description, and approving one of several record updates shows it as waiting until the others are answered
Layout and readability: The large text size keeps header icons in place, long cloud and model names stay within their space, and the device list and delete-account dialog scroll on phones

📜 Legal

Privacy Policy and Terms: Both now cover local models, Ollama, and model servers you connect from the desktop app

Release Date: October 8th, 2026

Version: 2.6.0

Desktop Version: 1.0.5