# Explore Chat
Source: https://help.heymumble.com/guides/ai-chat/explore-chat



**Use the Chat in the Way That Works Best for You**

You can start a voice comment at any moment with the shortcut key. This reduces context switching and keeps your workflow smooth and uninterrupted.

More voice skills will continue to roll out, helping you think faster and stay productive with less effort.

**Message Input Toolbar**

The message bar gives you flexible ways to share content with Mumble. It supports quick uploads and helps Mumble understand more context about your work.

You can use the toolbar to:

* drag and drop images into the chat
* attach local files from your device


# How to Access Mumble
Source: https://help.heymumble.com/guides/ai-chat/how-to-access-mumble



Mumble is your voice first AI agent. It helps you capture ideas, generate drafts, and manage your calendar and notes.

**Where to Find Mumble**

You can get access to Mumble from Home page or the bottom right corner of the Workspace page. Mumble also appears automatically inside key features whenever it can help you complete a task.

**You will see Mumble when you**

* create templates in Notes
* edit smart note rules

**How to Start a Chat with Mumble**

You can talk to Mumble using your voice or your keyboard. Both methods work the same and Mumble reacts instantly.

* Hold the **"fn"** key and speak to begin a voice chat. Mumble listens and responds based on what you say
* Type directly into the chat input if you prefer text

**AI Chat and processing mode**

AI Chat runs on cloud models today and requires Cloud Mode. See [Cloud Mode vs Local Mode](/guides/settings/cloud-vs-local-mode) for what's available on-device.


# What Mumble Can Help With
Source: https://help.heymumble.com/guides/ai-chat/what-mumble-can-help-with



Mumble is more than a voice dictation tool. It understands what you want to do, generates content, manages your calendar and contacts, and helps you stay organized while thinking faster across your workspace.

Mumble can:

* manage tasks, contacts, and notes directly inside the app
* draft content based on your instructions
* answer questions to help you plan, brainstorm, and clarify ideas
* turn your voice comments into structured outputs whenever you need them

Mumble grows with your workflow and is always ready to help you get things done!


# Customize Your Dictation Output
Source: https://help.heymumble.com/guides/dictation/customize-dictation-output



When you start dictating with Mumble by holding down your dictation shortcut (e.g. Fn), a small panel appears showing a row of format shortcuts.

<img alt="Dictation panel" />

While still holding the dictation key, tap any of these shortcut keys to change how your text will be formatted. For example, ← for Concise, ↑ for Bullets, or / for AI Skills. Release the dictation key when you're done, and the formatted result is inserted.

You can press a format shortcut at any point while dictating (before, during, or after speaking), as long as you're still holding the dictation key.

## How the panel works

1. Hold your dictation shortcut (Fn by default, customizable in **Settings > Shortcut**) and start speaking.
2. While still holding, press a key to choose an output format (optional).
3. Release to insert the result at your cursor.

If you don't press any key, Mumble uses the default action (← Concise).

## Output actions

While holding your dictation shortcut, press one of these keys to pick how your speech is processed. The selected action highlights in the panel.

**← Left arrow: Concise.** Polishes your speech into clean text. Fixes grammar, punctuation, and formatting. This is the default.

<Frame>
  <img alt="Dictation Default" />
</Frame>

**↑ Up arrow: Bullet points.** Formats your speech as a bullet point list.

<Frame>
  <img alt="Dictation Bullet" />
</Frame>

**. Period: Open.** Extracts a URL from your speech and opens it in the browser.

<Frame>
  <img alt="Dictation Open" />
</Frame>

**→ Right arrow: Custom prompt.** Applies your active custom prompt to transform the text. You need to set this up first (see below).

<Frame>
  <img alt="Dictation Custom" />
</Frame>

**↓ Down arrow: Emoji.** Cancels dictation and opens the macOS emoji keyboard.

**, Comma: Insert a GIF.** Pick a GIF and insert it at your cursor.

**/ Slash: Run a skill.** Invokes a skill on your dictation. See the Skills section for details on what skills are available and how to set them up.

**Esc: Cancel.** Cancels dictation without inserting anything.

## Customize your dictation panel

You can choose which hotkeys appear in your dictation panel, so only the ones you actually use show up.

1. Click **Dictation** in the sidebar.
2. Open the **Quick settings** tab.
3. Find the **Hotkey panel** section.
4. Toggle on the hotkeys you want to show, and toggle off any you don't need.

The Custom prompt row works differently. Instead of a toggle, it has a dropdown to choose which of your saved prompts is active for the **→** action. Learn how to create one in the [Set up a custom prompt](#set-up-a-custom-prompt) section below.

If you don't want the panel to show at all, click **Disable all** at the top of the section. The panel won't appear during dictation.

<Frame>
  <img alt="Dictation quick settings" />
</Frame>

## Set up a custom prompt

The → key triggers whatever custom prompt is active. To set up a custom prompt:

1. Click **Dictation** in the sidebar.
2. Go to the **Custom prompts** tab.
3. Click **+ New prompt** and enter your prompt.
4. Toggle it on when you're done.

During dictation, press **→** and your output will follow that custom prompt.

<Frame>
  <img alt="Custom prompts tab" />
</Frame>

## Custom prompt ideas

Your prompt can be anything that transforms text. A few examples:

* **Translate.** "Translate into Spanish." Dictate in English and get Spanish output.
* **Change tone.** "Rewrite in a formal, professional tone." Turn casual speech into polished writing.
* **Shorten.** "Condense to one paragraph." Get a tight summary of a longer thought.
* **Reformat.** "Rewrite as a numbered step-by-step guide." Turn spoken instructions into structured steps.


# Dictation Overview
Source: https://help.heymumble.com/guides/dictation/dictation-overview



Global Dictation lets you type by voice in any app on your Mac. Hold your dictation shortcut, speak, and Mumble inserts clean text at your cursor. It works in your email client, Slack, Google Docs, your code editor, or anywhere you can type.

## Turn it on

Global Dictation needs to be enabled in **Settings > Dictation > Global Dictation**. Toggle it on and you're ready to go.

<Frame>
  <img alt="Global Dictation toggle" />
</Frame>

## How to dictate

Hold your dictation shortcut (default is Fn) and start speaking. Mumble transcribes your speech in real time. When you release the key, the text is inserted at your cursor. If Mumble doesn't detect an active text field, a floating clipboard appears with your transcribed text. Click the copy button to paste it anywhere.

You can customize the dictation shortcut in **Settings > Shortcut**. The dictation shortcut must be different from the AI skill shortcuts.

## Dictation and Local Mode

When Local Mode is enabled, dictation is fully local. Everything from speech recognition to text output runs entirely on your Mac. No audio or text is sent anywhere. This also means dictation works offline. [Learn more →](/guides/settings/cloud-vs-local-mode)

## Language

Mumble uses the language you selected during setup to optimize transcription accuracy. You can change it anytime in **Settings > Preference > Languages**. Local Mode works best with a single language selected. Multiple languages may reduce accuracy.

## Dictation vs voice notes

Dictation is purely for typing by voice. It inserts text into whatever app you're using. Voice notes are different. They are triggered from the AI Skills panel (default double tap Fn), always return to Mumble, and can record audio, create notes, or perform actions. [Learn more about voice notes →](/guides/notes/notes-overview)


# Coming Soon
Source: https://help.heymumble.com/guides/iphone/coming-soon



Mumble AI for iPhone is currently in development. Stay tuned for updates.

We're bringing the same powerful voice-first experience to your iPhone, so you can capture ideas, meetings, and notes on the go.

**What to expect:**

* Voice notes on mobile
* Meeting summaries synced with your Mac
* AI Chat accessible anywhere
* Voice skills triggered from your phone
* Push notifications for upcoming meetings

Want to be notified when it launches? Reach out to us at [feedback@mumblenote.com](mailto:feedback@mumblenote.com).


# During a Meeting
Source: https://help.heymumble.com/guides/meetings/during-a-meeting



Once recording starts, Mumble opens a floating panel where you can see live transcript, take notes, and manage attendees without leaving your meeting.

## The meeting panel

Once recording starts, Mumble opens a floating panel that keeps two things in view while you talk: a space for your own notes, and a live transcript of the conversation. From the panel you can also check the meeting duration, pause the recording, manage attendees, or finish and delete the meeting.

<Frame>
  <img alt="Meeting panel" />
</Frame>

## Live transcript

Click **"Live transcript"** at the bottom of the panel to see the conversation transcribed in real time. Each line is marked as either Microphone (your voice) or System Audio (other participants). Mumble then works in the background to identify who is speaking. Initially, speakers may appear as Speaker 1, Speaker 2, and so on. As the meeting progresses, Mumble reads participant names from the meeting window and assigns them to the correct speakers. This requires Screen Recording permission.

If a speaker isn't identified automatically, you can click **"+"** on the attendee row to add the person manually.

## Take your own notes

The Notes section in the panel lets you type your own thoughts while the meeting is happening. Use it to flag key moments, capture decisions, or write follow-ups as you need. Anything you write here will be included in the meeting summary after the recording ends.

## Stay focused

If the panel gets in the way, click the shrink button (top right) to minimize it into a small floating widget. Or click the hide button (the eye icon next to it) to tuck Mumble into the menu bar, where it keeps recording silently. Click the Mumble icon in the menu bar to bring it back.

When the meeting is over, click the send button to end the recording and generate your summary. If [Auto Recording](/guides/meetings/meetings-overview#auto-recording) is on, Mumble detects that the meeting has ended, shows a 10-second countdown, and automatically saves the recording and starts generating your summary. [Learn more about meeting summaries →](/guides/meetings/meeting-summary)


# Meeting Summary
Source: https://help.heymumble.com/guides/meetings/meeting-summary



After a meeting ends, Mumble helps you turn the conversation into a clear and useful summary. You can choose different templates, regenerate outputs, and export the results to share with others.

<iframe title="Mumble AI Meeting Summary" />

* Meeting summary template

  Inside the summary block, click **"Auto"** to open the template menu. You can select different templates to structure your meeting summary.

* Meeting output

  You can also choose different output templates to regenerate the meeting content. When you click **"+ New output"** and apply a new output template, the newly generated content appears below the existing summary. The original content is not overwritten, so you can compare different versions and keep what works best for you.

* Regenerate and export

  If you are not satisfied with the summary, click **"Regenerate"** in the top right corner to create a new version.

  To share or save your summary, click **"Export to Google Docs"**. This makes it easy to send meeting notes to teammates or store them with your other documents.


# Record a Meeting
Source: https://help.heymumble.com/guides/meetings/meetings-overview



Mumble records meetings directly from your Mac. No bot joins the call. It works with Google Meet, Zoom, Teams, Slack, and any other platform that plays audio through your Mac.

<iframe title="Mumble AI Meeting Recording" />

## Ways to start recording

There are several ways to start a meeting recording in Mumble.

### From a calendar reminder

When you connect Google Calendar, Mumble sends you a system notification one minute before each meeting. The notification shows the meeting title, time, and participants. Click **"Join meeting"** to open the call and start recording.

<Frame>
  <img alt="Calendar reminder notification" />
</Frame>

### From automatic detection

When Mumble detects a meeting app like Zoom, Google Meet, or Teams is active, a notification appears saying "Meeting Detected" with a countdown. Click **"Record"** to start, or **"Cancel"** to skip. If Auto Recording is enabled, Mumble starts recording automatically without waiting for you to click.

<Frame>
  <img alt="Meeting detected notification" />
</Frame>

### Record a new meeting

Click **"New meeting"** from the Home screen, the sidebar, or the top of the Meetings page to start a recording manually. This is useful for ad-hoc calls or in-person conversations that aren't on your calendar.

<Frame>
  <img alt="Record a new meeting" />
</Frame>

### From a calendar event

Click any upcoming meeting synced from Google Calendar to open its detail page, then click **"Start recording."** This links the recording to that calendar event so the meeting title and participants carry over automatically.

<Frame>
  <img
    alt="Calendar event detail page with start
recording"
  />
</Frame>

## Auto Recording

Auto Recording is on by default. When enabled, Mumble automatically starts and stops recording when meetings begin and finish, with a 10-second countdown before it begins. Turn it on in **Settings > Meeting > Auto Recording**.

<Frame>
  <img alt="Auto recording setting" />
</Frame>

## How recording works

Mumble captures your microphone audio and system audio at the same time. This means it records both your voice and everything the other participants say through your speakers or headphones. No bot or extra participant joins the call.

Meeting recording and transcription are available in both Cloud Mode and [Local Mode](/guides/settings/about-local-mode). In Local Mode, the entire pipeline runs on your Mac with no audio or transcript leaving your device.

## Create upcoming meeting

Click the **"+"** next to Meetings in the sidebar and choose **"New upcoming meeting"** to create one before it starts. Set the title, time, and participants. From the **"..."** menu, you can choose a template for the meeting summary and sync the event to Google Calendar.

When the meeting begins, Mumble will notify you to start recording. You can also open the meeting anytime and click **"Start recording"** manually.


# Install & Set Up
Source: https://help.heymumble.com/guides/mumble-ai-101/getting-started



Get Mumble AI installed on your Mac and walk through the onboarding steps.

## System requirements

Mumble AI requires macOS 14.2 or later and Apple Silicon (M1 or above). Intel Macs are not supported yet.

If you plan to use Local Mode, Apple Silicon with 24GB or more memory is recommended.

## Install Mumble AI

1. Download Mumble AI from [heymumble.com](https://www.heymumble.com)
2. Open the downloaded file and drag Mumble AI to your Applications folder
3. Launch the app

<Frame>
  <img alt="Drag Mumble to Applications" />
</Frame>

## Sign in with Google

Mumble uses Google sign-in. You can log in with either a Google Workspace account or a personal Gmail account.

<Frame>
  <img alt="Continue with Google" />
</Frame>

Click "Continue with Google" and a browser tab will open for you to complete authentication. If you have multiple Google accounts, pick the one tied to your work calendar. Mumble pulls in your calendar events to power meeting detection and auto-recording.

<Frame>
  <img alt="Link primary Google account" />
</Frame>

## Connect your Google Calendar

During sign-in, Mumble will ask for permission to access your Google Calendar. Grant this to get the full experience. With calendar access, Mumble can show your upcoming meetings in the app, remind you before a meeting starts, automatically start and stop recording when a meeting is in progress, and let you join calls with one click.

<Frame>
  <img alt="Grant calendar permission" />
</Frame>

You can skip this step and still use Mumble, but meeting features will be limited. To connect later, go to **Settings > Integrations > Google Accounts**, or click the prompt on your Home screen > upcoming meetings.

## Choose your processing mode

Back in Mumble after sign-in, Mumble starts by asking what you plan to use it for: Meeting Notes, Voice Notes, or Dictation. Pick all that apply. Then choose how Mumble should process your audio.

* **Cloud.** Best accuracy and fast performance, powered by the latest large language models. Requires internet. Audio is processed in the cloud and never stored.
* **Local.** Full privacy. Audio never leaves your Mac and works offline. Apple Silicon with 24GB+ memory is recommended.
  * If your Mac meets the requirement, Mumble starts downloading the local model right away.
  * If your device has less than 24GB, Mumble will let you know the local model may run slowly or use a lot of memory. You can still proceed with the download or switch to Cloud.

You can change this anytime in Settings. [Learn more →](/guides/settings/cloud-vs-local-mode)

<Frame>
  <img alt="Choose processing mode" />
</Frame>

## Grant permissions

Mumble then walks you through a few macOS permissions it needs before you can get started. Which permissions you see depends on the features you selected in the previous step.

The first two are required for Mumble to function at all:

* **Microphone Access.** Lets Mumble hear your voice for recordings, voice notes, and dictation.
* **System Audio Recording.** Lets Mumble capture audio from other apps, so it can hear the other participants in your meetings.

The next two are optional but recommended. Without them, you can still record meetings and dictate, but you'll miss out on advanced features like auto-recording, speaker identification and dictation enhancement:

* **Accessibility.** Enables Mumble to detect when meetings start and end, so recording can begin and stop automatically. Lets Mumble read the context in the active text field, so it can produce better output for dictation.
* **Screen Recording.** Enables Mumble to read participant names from the meeting window, so your transcript gets accurate speaker labels.

You can review and update these anytime in **Settings > Permissions**, or in **macOS System Settings > Privacy & Security**.

<Frame>
  <img alt="Set up permissions" />
</Frame>

## Set up your shortcuts

Mumble uses two keyboard shortcuts. During setup, you'll configure them and make sure they work on your Mac.

* **Hold Fn** for dictation and AI chat.
* **Press Fn + Space** for voice notes and skills.

These are the defaults. If they conflict with other apps or just don't feel right, click "No, change shortcuts" to customize them. You can also change shortcuts later in **Settings > Shortcut**.

<Frame>
  <img alt="Test your shortcut" />
</Frame>

## Choose your language

Select the language you use most often. Mumble uses this to optimize transcription for meetings, voice notes and dictation. You can change it later in **Settings > Preference > Languages**.

<Frame>
  <img alt="Choose your language" />
</Frame>

## Try each feature

Based on your earlier selections, Mumble walks you through a quick interactive demo of each feature you chose. You'll try dictation, create a voice note, or see how Mumble records a meeting and generates a structured summary, using the shortcuts you just configured.

## You're all set

After onboarding, you land on the Home screen. If you connected Google Calendar, your upcoming meetings are already synced and ready to go.


# Home & Navigation
Source: https://help.heymumble.com/guides/mumble-ai-101/homepage-overview



Home is your command center in Mumble. Every key action is accessible from this screen: start a new meeting, create a voice note, use a skill, or chat with AI. It also serves as a dashboard where you can track what Mumble has done for you, what's coming up, and what still needs your attention. The sidebar on the left gives you access to every other part of the app.

## Getting started checklist

When you first open Mumble, a checklist appears at the top of Home to help you try each core feature. Each card walks you through one action.

* **Capture an idea.** Create your first voice note and see how Mumble turns it into a clean, structured note.
* **Start a meeting.** Record a meeting and see live transcripts and a summary generated afterward.
* **Schedule a Google Calendar event.** Try creating a calendar event by voice, like "Lunch with Alex tomorrow at 12."

<Frame>
  <img alt="Getting started checklist" />
</Frame>

As you complete each one, the card gets checked off. You can also dismiss items manually. Once you've finished or cleared them all, the checklist disappears and won't come back.

## Dashboard

Below the checklist (or at the top of Home once it's gone), you'll see two columns that act as your dashboard.

* **On Track** shows items that are fully processed and complete. This includes meeting notes with finished transcripts and summaries, voice notes that have been structured, and completed skill actions. The number next to the heading is your total count. A blue dot on a card means it's new since you last checked.
* **Needs your input** shows items that still need something from you. This could be a skill action that requires confirmation. When nothing is pending, this column shows "Nothing yet."

<Frame>
  <img alt="Dashboard" />
</Frame>

## Filters

The controls at the top right let you switch between different views of your Home content.

* **Update time** sorts items by when they were last modified. You can change it to Create time.
* **Upcoming meeting** opens a panel showing your schedule synced from Google Calendar, organized by day with meeting times, titles, and participants.
* **Archived** shows items you've previously archived.

<Frame>
  <img alt="Filters" />
</Frame>

## App sidebar

The left sidebar is how you navigate each core function of Mumble.

* **Home.** Returns you to this screen.
* **Notes.** All your text notes and voice notes. Create, edit, search, and organize them here. [Learn more →](/guides/notes/notes-overview)
* **Tasks.** A space for your to-dos. When you create a voice note and choose "task note" as the type, the content goes here automatically. You can also type tasks directly or record them by voice.
* **Meetings.** Every meeting Mumble has recorded, with audio, transcripts and summaries. You can also create upcoming meetings manually or see ones synced from Google Calendar. [Learn more →](/guides/meetings/meetings-overview)
* **People.** Your contact list, built automatically from meeting participants or people mentioned in your chats with Mumble AI. View all meetings and notes associated with a specific person.
* **Dictation.** Dictation settings and history. This feature is currently in alpha, so it may not be visible in your app yet.

<Frame>
  <img alt="Sidebar" />
</Frame>

## Resources

Under a collapsible section in the sidebar.

* **Dictionary.** Add custom vocabulary so Mumble recognizes names, acronyms, or technical terms that standard transcription might miss.
* **Templates.** Browse and manage templates that control how Mumble formats meeting summaries and voice notes. [Learn more →](/guides/template/template-overview)
* **Tutorial.** A series of video demos for a refresher on each feature.

## Inbox and other controls

At the bottom of the sidebar.

* **Inbox.** Notifications and updates from the Mumble team.
* **Feedback.** Send feedback directly to the Mumble team.
* **Check for Updates.** Make sure you're running the latest version.


# What is Mumble AI
Source: https://help.heymumble.com/guides/mumble-ai-101/what-is-mumble-ai



Mumble AI is a voice-first productivity app for Mac. It records your meetings without adding a bot to the call, turns your voice into clean notes, and lets you dictate text in any app on your Mac.

## What Mumble does

* **AI meeting assistant, no bot joins the call.** Mumble captures audio directly from your Mac. It works with Google Meet, Zoom, Teams, Slack, and other meeting platforms. During the meeting, you get a live transcript with speaker labels. After the meeting, Mumble generates a full meeting summary. [Learn more →](/guides/meetings/meetings-overview)

* **Capture ideas with your voice.** Start a voice note and Mumble records your audio, transcribes it, and can restructure the content using templates, so a rambling two-minute thought becomes a clean outline, a to-do list, or whatever format you need. [Learn more →](/guides/notes/notes-overview)

* **Dictate text anywhere on your Mac.** Hold Fn or your custom shortcut to dictate directly into any app. Mumble transcribes your speech and inserts clean, polished text at your cursor. You can take this further with custom actions that reformat the output, translate it into a different language, adjust the tone, or change the length. [Learn more →](/guides/dictation/dictation-overview)

## Voice Skills <span>Coming soon</span>

Mumble is expanding beyond capture into actions. Voice Skills let you trigger tasks by speaking, like scheduling a calendar event or sending a message through Slack or WhatsApp. You'll be able to import any skill already available on your Mac from other agents, or explore recommended skills in Mumble's skill library.

## Cloud Mode and Local Mode

Mumble offers two ways to process your audio. Cloud Mode sends audio to the cloud for the best speed and accuracy. Local Mode runs all processing on your Mac for complete privacy. You can choose which mode to use and set it independently for meetings, voice notes, and dictation. [Learn more →](/guides/settings/cloud-vs-local-mode)

## Mumble Note for iPhone

[Mumble Note](https://www.mumblenote.com) is a companion iOS app for capturing voice notes on the go. Notes you create in Mumble Note one-way sync into your Mumble AI workspace on Mac. You capture on your phone, and everything shows up in Mumble AI for you to organize, search, and build on.


# AI Notes
Source: https://help.heymumble.com/guides/notes/ai-notes



AI Note reads everything in your note and generates a clean, structured version of it using a template you choose. The original note stays untouched and your AI Note is created as a separate note, linked back to the original as its source.

To create an AI Note:

1. Open any note.
2. Click the **+** button in the top right.
3. Under **Create a new AI Note**, pick a template from the dropdown, or use one you've created yourself. See [Create a Template](/guides/template/creating-template) to set one up.
4. Mumble generates a new AI Note based on your original content and the template you chose.

<Frame>
  <img alt="Create a new AI Note" />
</Frame>

## Keeping your AI Note up to date

Once an AI Note is created, it automatically updates 20 minutes after you make changes to the source note. To refresh it immediately:

1. Open your AI Note.
2. Click the **AI Note icon** in the top right to open the AI Note panel.
3. Click **Regenerate now**.

Your AI Note updates instantly with your latest changes.

<Frame>
  <img alt="Regenerate AI Note" />
</Frame>

## AI Notes and Local Mode

AI Notes can be generated fully on-device with Quick Format templates when [Local Mode](/guides/settings/about-local-mode) is enabled. Customized templates currently require Cloud Mode.


# Notes Overview
Source: https://help.heymumble.com/guides/notes/notes-overview



Notes are the core space in Mumble where you capture, organize, and revisit your thoughts. Whether you type, speak, or generate content with AI, everything lives in notes so it is easy to manage and edit later.

* Layout

  In the Notes view, the notes list appears on the left and the selected note opens on the right. At the top of the list, you can use filters to decide how notes are displayed, either sorted by last updated time or by creation time.

  If you click the 🌟 icon in the top right corner of a note, it will be added to your favorites. You can quickly find all your starred notes in the **Favorites** section in the left sidebar or from the **Favorites** tab at the top of the Notes page.

  When you create Smart Note rules and turn **“Smart Notes”** on, notes that match those rules will be automatically organized. You can view them anytime in **Smart Notes** tab at the top of the Notes page.

* Note types: text note, voice note

  Mumble supports two main types of notes: text notes and voice notes. You can edit notes by typing directly, or by clicking the **"+"** on the left to add different blocks.

  * For quick capture, hold **"fn"** to start voice dictation. What you say will be transcribed into text and appear directly in your note.
  * If you want to keep the original recording, press **"option" + "space"** to create a voice note block.

  The difference between voice dictation and voice notes is dictation only turns speech into text, but voice notes keep the original audio and transcript, and let you apply templates or regenerate structured content.

  Voice notes can be captured and transcribed fully on-device in [Local Mode](/guides/settings/about-local-mode).

* Basic actions
  * Create a note Click the “+” next to Notes and select New note. You will see a blank note. The note is saved only after you add a title or content.
  * Create a voice note Click the “+” next to Notes and select New voice note. This opens the same voice note panel as pressing option + space.
  * Voice edit Select the part of the note you want to change, hold option, and say what you want it to become. Mumble will update the text for you.


# Smart Notes
Source: https://help.heymumble.com/guides/notes/smart-notes



Smart notes help you automatically organize messy notes into structured ones. You set the rules once, and Mumble takes care of the rest.

* Smart notes are designed to:
  * Automatically structure and organize your notes
  * Continuously collect related voice notes into a single organized note
* How to use
  * Set rules

    You can set rules in two ways:

    * Ask Mumble to generate rules for you
    * Create the rules yourself

  * Edit rules

    You have three ways to edit Rules:

    * You can click **"Edit with AI"** and describe the changes you want
    * You can manually add or remove rule details
    * You can select part of a rule and use voice edit by holding option


# About Local Mode
Source: https://help.heymumble.com/guides/settings/about-local-mode



Local Mode is Mumble running entirely on your Mac. No audio is sent anywhere. No transcripts are sent anywhere. The model that turns your voice into notes runs on your device.

This page covers what Local Mode does, how to set it up, and what to expect.

## What Local Mode covers

Local Mode is a complete offline pipeline for Mumble's core features:

* **Meetings.** Record any conversation on your Mac with no bot joining the call. Transcription, speaker separation, and AI Notes all run on-device.
* **Voice Notes.** Capture a thought, get a clean transcript and structured note without an internet connection.
* **Dictation.** Speak anywhere on your Mac. Text appears in the active app, processed locally end to end.
* **[AI Notes](/guides/notes/ai-notes) and Quick Format templates.** Apply Quick Format templates and generate AI Notes from any local recording, fully on-device.

[Customized templates](/guides/template/template-overview), [Voice Skills](/guides/skills/skills-overview), and [AI Chat](/guides/ai-chat/how-to-access-mumble) are not available in Local Mode today.

## What "only you can read it" means

In Local Mode:

* Audio is captured and processed on your Mac.
* Transcripts stay on your Mac.
* AI Notes and Quick Format output are generated on your Mac.
* Nothing is sent to Mumble servers or any third party for processing.

No bot, no cloud upload, no transcript leaving your device. Your recordings are readable only by you, on your machine.

## Requirements

Local Mode needs enough on-device compute to run the models well. Current requirements:

* Apple Silicon Mac (M-series). Intel Macs are not supported.
* macOS 15 or later.
* 24GB unified memory or more recommended for the best experience. Lower configurations may work but can be slower on long recordings.
* Storage for the local models. The download happens once.

## Setting it up

1. Open **Settings > Processing Mode**.
2. For each feature you want to run locally (Meetings, Voice Notes, Dictation), switch the toggle to **Local**.
3. If this is your first time, Mumble will prompt you to download the local models. Keep the app open until the download finishes.
4. Once installed, Local Mode works without internet.

You can also choose Local Mode during onboarding when Mumble first sets up.

Prefer the fastest results and broadest language coverage? See [Cloud Mode vs Local Mode](/guides/settings/cloud-vs-local-mode) to compare.

## When to use Local Mode

Local Mode is the right choice when:

* The conversation is confidential and you want a clear privacy boundary.
* You work in a regulated environment (legal, medical, finance, internal corporate matters).
* You're on a flight, in a secure facility, or anywhere without reliable internet.
* You simply prefer your voice data to stay on your machine.

For most everyday meetings and notes, Cloud Mode is faster and supports 40+ languages out of the box. Many users run Cloud for general work and Local for sensitive sessions. You can switch per feature anytime in **Settings > Processing Mode**.

## Language support

Cloud Mode supports 40+ languages with automatic detection.

Local Mode supports multiple languages, but the local dictation model works best when you select a single language at a time. If you regularly switch between languages during dictation, Cloud Mode will give you better results.

## Free forever

Local Mode is a core part of what Mumble is. It will always be free.


# Cloud Mode vs Local Mode
Source: https://help.heymumble.com/guides/settings/cloud-vs-local-mode



Mumble gives you two ways to process your voice: Cloud Mode and Local Mode. Both are full-featured. You choose per feature, based on what each conversation needs.

## What's the difference

**Cloud Mode** runs on our servers. You get the fastest processing, the highest accuracy, and support for 40+ languages. Audio is encrypted in transit and at rest, and is never stored after processing.

**Local Mode** runs entirely on your Mac. Nothing leaves your device. Recording, transcription, AI Notes, and quick format templates all happen on-device. Works offline.

## What runs locally

Local Mode covers Mumble's core capture and notes:

| Feature                                                                                | Cloud | Local |
| -------------------------------------------------------------------------------------- | :---: | :---: |
| [Meetings](/guides/meetings/meetings-overview)<br />(record + transcript + AI summary) |   ✓   |   ✓   |
| [Voice Notes](/guides/notes/notes-overview)                                            |   ✓   |   ✓   |
| [Dictation](/guides/dictation/dictation-overview)                                      |   ✓   |   ✓   |
| [AI Notes](/guides/notes/ai-notes)                                                     |   ✓   |   ✓   |
| Quick Format templates                                                                 |   ✓   |   ✓   |
| Customized templates                                                                   |   ✓   |       |
| [Voice Skills](/guides/skills/skills-overview)                                         |   ✓   |       |
| [AI Chat](/guides/ai-chat/how-to-access-mumble)                                        |   ✓   |       |

For Meetings, Voice Notes, and Dictation, Local Mode handles the full pipeline on your Mac, from audio capture to structured notes. Customized templates, Voice Skills, and AI Chat are Cloud only today.

Meetings is the clearest example: recording, transcript, and AI summary all run on-device. You can take a meeting on a plane or in a secure room with no internet, and nothing about it (audio, transcript, or summary) ever leaves your Mac.

## How to choose

You don't have to pick one mode for everything. Each feature has its own toggle in **Settings > Processing Mode**.

A simple way to think about it:

* **Cloud** for everyday meetings and notes where you want the fastest results and broadest language coverage.
* **Local** for sensitive conversations, regulated work, or anytime you want a clear privacy line. Also useful when you're offline.

For a deeper look at what runs on-device and how to set it up, see [About Local Mode](/guides/settings/about-local-mode).

## Requirements for Local Mode

* Apple Silicon Mac (M-series)
* macOS 15 or later
* 24GB unified memory or more recommended
* One-time model download (about a few GB, done during onboarding or from **Settings > Processing Mode**)

A note on language: the local dictation model works best when you set a single language. Selecting multiple may reduce accuracy. Cloud Mode handles 40+ languages without this limitation.

## Set it per feature

Open **Settings > Processing Mode**. Each feature has its own Cloud and Local toggle. Switch anytime.

<Frame>
  <img alt="Cloud and Local Mode" />
</Frame>


# Skills Overview
Source: https://help.heymumble.com/guides/skills/skills-overview



Skills are ready-to-use AI workflows you can run in Mumble. Each skill is built for a specific task: brainstorming ideas, drafting a social media post, or running competitor research, and gives you a structured output instead of a free-form chat.

## Browse and install skills

Open the **Skills** tab in the sidebar to browse. You'll see two tabs:

* **Discover**: A curated collection of skills, including ones built by the Mumble team and open-source skills we've packaged for you. Click any skill card to see a brief introduction, its workflow, and example use cases.
* **My Skills**: All the skills you've installed.

<Frame>
  <img alt="Skills Discover tab" />
</Frame>

To install a skill:

1. Hover over a skill card in **Discover**.
2. Click **Install** in the top right.

The skill now appears in **My Skills** and is ready to use. See [Use a Skill](/guides/skills/use-a-skill) to learn how to trigger it.

To uninstall a skill, open its detail page and click **Delete**, or use the delete button on the skill card.

## Request a skill

If there's a skill you'd like but can't find, you can request it from the Mumble team.

1. Click **Request a skill** in the top right of the Skills page.
2. Paste a reference link (a GitHub repo, article, or docs page that describes the skill you want).
3. Describe what the skill should do. You can hold the shortcut shown in the field to dictate it.
4. Optionally, attach files for additional context.
5. Add your contact info so we can reach you when the skill is ready.
6. Click **Send**.

We'll notify you by email and in your Mumble inbox once the skill is added.

## Skills and processing mode

Voice Skills run on cloud models today and require Cloud Mode to be enabled. See [Cloud Mode vs Local Mode](/guides/settings/cloud-vs-local-mode) for the full breakdown of what runs where.


# Use a Skill
Source: https://help.heymumble.com/guides/skills/use-a-skill



There are two ways to use a skill in Mumble.

## From a skill's detail page

1. Open any skill from **My Skills** or **Discover**.
2. Click **Run** in the top right.
3. The Voice Launchpad opens with that skill selected. Describe your task by voice.
4. Press **Enter** or click the **Send** button.

The **Try it** button on a skill card works the same way.

## From the Voice Launchpad

Press your shortcut (double-press Fn by default) to open the Voice Launchpad from anywhere on your Mac. The launchpad opens with Voice Note selected by default.

To run a skill:

1. Pick the skill you want from the Skills list.
2. Describe your task by voice.
3. Press **Enter** or click the **Send** button.

<Frame>
  <img alt="Voice Launchpad" />
</Frame>

After you speak and send, the skill opens in a workspace with two sides. The right side is your interaction panel, where you'll see the agent's progress and reasoning. The agent may ask you questions along the way. Just click to answer and the skill continues. The left side works like a mini app, showing your input, the context the agent has gathered, and the structured output as it builds.

<Frame>
  <img alt="Running a skill" />
</Frame>

When the skill finishes, most text-based skills save the output as a note in your Notes, so you can come back to it anytime. Other skills take an action elsewhere. For example, the scheduling skill books an event in your Google Calendar.


# Create a Template
Source: https://help.heymumble.com/guides/template/creating-template



You can create a template in two different ways: with AI assistance or manually.

At the top of the template list on the left panel, you’ll see a **+ New Template** button. Click it to choose between: Generating a template with AI and Creating one yourself.

* Select **“Describe what you need”** and tell Mumble what this template is for. You can describe the purpose, structure, or sections you would like to include.
* If you choose **“Create manually”**, you can:
  * Type directly into the editor
  * Use voice dictation shortcut to input content by talking


# Manage Templates
Source: https://help.heymumble.com/guides/template/managing-template



Templates can be customized, previewed and prioritized based on your workflow.

**Editing**

All existing templates can be customized. You can modify their content, category, and order at any time.

* Category can be changed or added using the dropdown under the template title.
* The order of templates and categories can be adjusted in the left template list. You can drag and drop templates to reorder them or move them into a different category.
* Each section and the template instruction can be edited directly in the input fields on the right. You can also select specific text and use your shortcut for voice edit to quickly update the content.

**Preview**

If you want to see how a template works before using it, click Preview in the bottom right corner.

<Frame>
  <img alt="IMG 6502" />
</Frame>

In the preview page, you can generate output using either a built in sample from Mumble AI or your own notes. This helps you understand how the final result will look in real scenarios.

<Frame>
  <img alt="Group 292724" />
</Frame>

**Favorite templates**

You can click **Favorite** in the top right corner of a template to save it. Favorited templates will appear at the top of the template selection dropdown in meeting summary or notes.


# Templates Overview
Source: https://help.heymumble.com/guides/template/template-overview



Templates in Mumble are reusable structures used across multiple features, including [meeting summaries](/guides/meetings/meeting-summary), voice notes, and [AI Notes](/guides/notes/ai-notes) created from individual notes. You can create a template once and apply it wherever you need structured output.

**Template Type**

On the Template page, you will see two main categories at the top: Meeting Templates and Note Templates.

Meeting Templates are used for meeting summaries and meeting outputs. Note Templates are used for voice notes and text notes. Although they belong to different modules, these two types of templates support syncing. This gives you flexibility to reuse structures across different workflows.

**Layout: template list and overview**

The Template page is divided into two sections.

On the left side, you will see the template list. You can collapse categories you do not need, which helps keep your workspace clean and focused. On the right side, you will see the template overview.

<Frame>
  <img alt="Group 292714 Copy" />
</Frame>

When you select a template, the overview panel shows its instruction and included sections. Under the template title, you will see tags. “By Mumble” indicates a prebuilt template created by Mumble AI.