Voice memo

Last updated: January 27th, 2025

Overview

Voice memos in Tana mobile are a fast way to get your thoughts out of your head and into Tana for further processing. Your recordings are instantly transcribed on the server and can be automatically structured with the guidance of supertags and fields. The new iOS app (and soon Android!) provides instant transcription, supports 91 languages, and maintains easy access to original recordings through source material.

When a voice memo is recorded, it gets sent to your current daily note and is instantly transcribed and enriched by entities in your Tana.

You can further process voice memos using Rewrite, an AI feature for rewriting transcripts into other types of content.

Basics

• Quick start: In Tana mobile, go to Capture and start a voice memo. Speak, then stop it when you're done - your memo will be automatically transcribed and sent to the default Destination.
• While recording:
  • Select supertag to add structure: Tap "Supertag" to select a Supertag and to see its fields. Mention the field names explicitly for better auto-fill (e.g., if your field is called "Next steps", say "Next steps... send the proposal to client for review")
  • Change target Destination: Tap "Destination" before recording to send to a different location. Pick from Today, Inbox, and pinned nodes.
  • Select transcription language: Tana mobile will default to the system language of the device the first time it records a voice memo. You can choose another language from here.
  • Recording limit: 1 hour per memo
• Rewrite: Once the voice memo is transcribed in Tana, you can take the transcript and turn it into something else using AI. The default options include Summary, Pros and cons, Article and Tweet. If a custom rewrite command is present, it will hide the default ones. Supertags may offer custom rewrite options, tailored to that content.
• Access the original recording: Find a voice wave icon next to any voice memo node to access the original recording
• Only available on paid subscriptions: This feature requires AI and is not available on plans that do not include AI credits. It uses approximately 50 AI credits per 15 minutes of transcription

Details

How to record a voice memo on mobile

In the Tana mobile app, open the Capture tab and start a voice memo:

Once started, you will see the blue wave bars respond to what it hears:

While recording, you can do the following:

• Select a Supertag to use as a template (optional)
• Select a Destination (default shown as Today)
• Select a Language (bottom left, default shown as English)

Once done, complete the voice memo by hitting the blue send button, and it will be sent to the destination specified.

Select a supertag

Select a supertag to make its fields appear in the recorder interface.

• For supertags with many fields, scroll down to see the rest.
• Mention the field names for better auto-fill accuracy (e.g., "Highlight of the week... having lunch with an old friend")for Supertags with many fields, you can scroll down.
• Additional content will appear below auto-filled fields

Select a destination

Today is the target destination by default.

• To change that, long-press on the new target and confirm.
• Options for destinations include Today and Inbox, plus any pinned node.

Note for Inbox users

If you have set up your own command looking for audio files, or have tweaked the standard Inbox Audio Processing Command on the Inbox – this will still work, but you will not get the new voice memo processing.

Existing commands on the Inbox will continue to work as before, but be aware that these will require activity on the desktop client to run the processing, so the output will not be instantly available on mobile like with the new processing.

If you want to keep using the Inbox, you have two options (open link for instructions):

A) I want to keep using the Inbox as destination, but with the new processing

B) I want to keep using the Inbox with my own command/processing

Select a language

• Default: Uses your phone's system language
• Auto-detect: Works best for most common languages
• Manual selection: For best results with short memos in non-English languages, manually select the language. Choose from 91 supported languages.

Recording Limits

• Current limit: 1 hour per recording
• System notifies you when approaching the limit

Other ways to invoke voice recording on Tana mobile

There are several ways to initiate voice memos using Tana mobile:

• From the Capture tab: Press Capture, and select Voice memo
• From Supertag tab: Click the voice icon on a Supertag to start recording immediately.
• From any node: Press the blue plus button on the bottom right to get all the capture options
• From the iOS lock screen: Add the voice memo widget to your lock screen

Processing mobile voice memos

Once you have recorded a mobile voice memo, it will immediately be available on desktop and mobile:

• If no tag was selected: it will contain a cleaned-up transcription and will offer default options to Rewrite content.
• If a supertag was selected: Tana will run AI processes to fill in as much of its contents as possible, including title, description and fields
• Source material: Access the original recording and raw transcript via the voice wave icon present on the main node or node options menu.

Rewrite

You can find the Rewrite options on the top-right of the voice recording.

Pre-installed Supertags, or Supertags installed from Templates, may offer custom rewrite options tailored to that content, and the built-in Rewrite option will then not be shown.

The built-in rewrite offers these options, and will open a suggestion in a temporary draft panel:

Summary: Generates a summary of the transcript

Pros and cons: Generates a list of pros and cons. Tag the title with #article to publish and share immediately with Tana Publish

Article: Generates an article you can publish and share immediately with Tana Publish

Tweet: Generates drafts of short-form social media posts

How to adjust your voice memo workflows

AI instructions

In addition to using the supertag/field name and description as context for AI, Tana lets you add AI instructions on supertags or fields. This helps AI understand what the content is about, how you want the output formatted and give specific instructions. The AI instructions will be used when a supertag or field is included in autofill or mobile voice memo processing.

AI instructions support plain text, references and markdown. Common markdown formatting that may be useful:

• [ ] for checkbox
• # for Heading

Search nodes are not supported in AI instructions.

Open supertag/field configuration panel to edit the AI instructions:

Custom Autofill options

If you enable "Custom Autofill behavior" on a supertag, you can decide if you want Tana to try to autofill the node title, description and content/body. This includes the ability to set AI instructions specifically for each of these.

• Title - This could be used e.g. if you want to include the date in a certain format.
• Description - how you want the description of the node, e.g. "Make a one line summary of what the memo was about"
• Content - This could be used to specify a format you want for the body text instead of a basic transcript, e.g. "Make it a 140 character tweet" for a #tweet tag, or "Format it as an email".

To enable Custom Autofill behavior for voice memos, open the supertag configuration panel > AI and Commands.

Extract items

If you have a supertag that you use for voice memos, where you would like Tana to try to extract multiple tagged items from one recording, you can select a supertag in the Extract items field under Custom Autofill behavior. Tana will also try to autofill any fields on the tagged items it finds.

It's only possible to select one supertag to extract items for.

Exclude fields from autofill

If you have some fields you never want Tana to autofill, you can add them in the "Fields to exclude" section under Custom Autofill behavior. Open the field you want to exclude, copy the field, and paste it into the "Fields to exclude" section.

You can exclude a field by right-clicking and selecting: Disable autofill for field

Edit Rewrite prompts in Prompt editor

If you have AI commands that appear as command buttons for voice memos, you can iterate on the prompt directly in the draft panel by clicking "Run again":

By clicking Edit, you open the prompt from the underlying command, and can edit this to get the output you want. Changes will be saved to the underlying command when you click "Run again", and a new draft will be generated:

If you want to make more advanced prompts, referencing specific parts of the node or your graph, see Prompt expression keywords.

Known issues with mobile voice memos

• Auto-initialization of fields works only for date fields
• Search nodes not supported in AI instructions
• Multiple items from single recording currently limited
• Auto-fill can't be disabled

Desktop voice memos

There are two ways to record voice memos in Tana on desktop:

1. Live transcription

2. Record an audio file

Live transcription

See the notes in real-time as you talk, either for voice memos or in meetings.

There are several ways to start real-time transcription in Tana:

• Type / (slash) on an empty node, and choose "Live transcription" from the menu. The transcript will be shown below a new node where you can set a title.
• Create new button in sidebar > Start live transcription: Transcribes into the new node you created. New nodes like this are created on your Today page.
• Audio-enabled fields: Enables an audio button on the field. Pressing it triggers live transcription, with the transcript stored into the field. Set this up in field configuration.

Live transcription currently supports these languages:

• Chinese
• English
• French
• German
• Italian
• Korean
• Portuguese
• Spanish

Change live transcription language

To change the transcription language:

• Live transcription on a node: Click on the globe 🌐 icon
• Audio-enabled fields: Shift+click on the record button

Record an audio file

To start capturing a voice memo (no live transcription, no output shown until you stop the recording), you can type Ctrl/Cmd+K > Capture voice memo. A recording indicator will be shown in the sidebar, and you can click this to stop the recording.

If you're using the Tana Desktop app, you can do Cmd/Ctrl+Shift+E: This starts global live transcription that runs in the background. A floating recording menu will appear, and lets you stop the recording:

All recordings will be sent to the Inbox, which you'll see in the sidebar.

If your voice memo is not automatically transcribed, you can put your cursor on the title and do Ctrl/Cmd+K >Transcribe audio.

Unless you specified a default language to use during onboarding, your language is likely set to auto-detect.

If you're not having good results with this, you can specify a preferred language:

• Command line: Set default transcription language to → or Transcribe audio as.. and select a language

Desktop voice memos support more than 90 languages.