Changelog

Follow up on the latest improvements and updates.

RSS

new

macOS

iOS

Android

Web

Windows

Speaker Names in Your Transcripts (v3.9)

Hedy 3.8 made speaker labels accurate. Now they get names.
Hedy names them for you
  • When someone says who they are, or is addressed by name, the recap picks it up and applies it to the transcript.
  • Only actual dialogue counts as evidence, so ambiguous speakers stay numbered rather than guessed at.
Or you name them yourself
  • Each transcript shows a roster of speaker chips above it. Tap one to rename it.
  • A name you set is never overwritten by Hedy's own inference.
The name follows the session
  • Copies, shares, downloads and exports all use it.
  • It syncs to your other devices, and to sessions you have shared.
  • Transcript cleanup uses your confirmed names, so a tidied transcript reads with real people in it.
Names come from Nemotron's speaker analysis, so they appear on sessions captured with the
Nemotron
speech engine: Apple Silicon Macs, iPhone 12 and newer, iPad Air 4 and newer, Windows, and 64-bit Android phones. Sessions captured with Whisper or a cloud provider do not carry speaker turns and show no roster. Renaming is available on your own finished transcripts.

new

macOS

iOS

Android

Web

Windows

Consolidated Topic To-Dos (v3.9)

Every action item from a Topic, in one place.
  • A
    To-Dos
    tab on each Topic gathers the action items from all of that Topic's sessions.
  • Sort by due date or by the session they came from.
  • Edit, complete, delete and export them from the tab, and show or hide completed items.
  • Completing one there updates it in its original session.
Open a Topic and choose
To-Dos
. Topics shared with you show the list read-only.
Take a session out of Hedy as a real file.
  • Download the
    summary
    , your
    detailed notes
    , or a
    full session export
    as plain text, Markdown, Word (.docx) or PDF.
  • Word and PDF keep headings, lists, emphasis, quotes and tables instead of flattening them.
  • A full export gathers the recap, detailed notes, session notes, highlights, chat history and to-dos in one document.
Open a session and use
Download Summary
,
Download Detailed Notes
or
Download All Data
, then pick a format. Transcripts have their own
Download Transcript
action and save as a text file. PDF covers the characters in Hedy's bundled font; for anything outside it, Hedy points you to Word.

improved

macOS

iOS

Android

Web

Windows

Session Merges Keep Your Chat and Notes (v3.9)

Merging two sessions now carries the whole session across.
  • Post-session chat history travels with the merge, in order.
  • Session notes come along too.
  • If the chat history cannot be read reliably, the merge stops and tells you, leaving both original sessions untouched rather than producing a half-copied result.
Automatic suggestion popups are now adjustable without leaving them.
  • An in-card menu lets you hide alerts, change how often suggestions arrive, and choose whether popups follow your cursor or stay with the Hedy window.
  • Re-picking the frequency you are already on no longer resets your mid-session pause.
The matching display control appears under Meeting Settings, Show suggestions as alerts, only while alerts are switched on.
The API and MCP server handle large Topics properly.
  • GetTopicSessions accepts limit and startAfter, so a Topic with more than 50 sessions can be paged through instead of stopping at the first page.
  • Topic colours are accepted with or without a leading hash, matching what the app writes and the API returns.
  • The OpenAPI spec now documents session duration as whole minutes, and iconName as a numeric codepoint.
Requires API access and Hedy Pro.

improved

PRO

macOS

iOS

Android

Web

Windows

Accurate Plan Status in Your Account (v3.9)

Your Account screen tells you exactly where your subscription stands.
  • A Pro plan that is ending now says when access ends, instead of describing itself as renewing.
  • Renewing, ending and not-yet-known states are shown distinctly.
Under
Settings → Account → Your plan
. Lifetime plans are unchanged.
More choice when you bring your own speech provider key.
  • Deepgram adds Swiss German, Armenian, Gujarati, Nepali and Punjabi, plus more regional variants of Dutch, French, Portuguese, Spanish and Chinese.
  • OpenAI live sessions move to gpt-live-transcribe, and imported files to gpt-transcribe.
  • Existing saved selections keep working.
Find them under
Settings → Speech & AI → Speech Recognition Options
. Both providers need your own API key and are not available on the web app.
Your custom vocabulary reaches imported files, not just live sessions.
  • Names, jargon and acronyms you have added are used when transcribing an imported recording with Whisper.
  • On Android, that vocabulary now genuinely reaches the engine, so priming works there too.
Applies to the Whisper engine. Other speech providers use their own vocabulary settings.
Hedy is more precise about what it is hearing.
  • Barely-audible input and stopped audio are now told apart, so the warning matches the actual problem.
  • The wording no longer assumes a broken microphone, because the cause is often a muted app or a device that switched away.
  • The level indicator flattens when capture is unhealthy, so you can see the problem at a glance.
  • Tapping the warning opens Microphone Settings directly.
Load More