Hedy AI
Roadmap
Feedback
Changelog

Changelog

Follow up on the latest improvements and updates.

RSS

new

macOS

iOS

Android

Web

Windows

Transcript Timestamps (v3.12)

See when each part of the conversation was said, and jump straight to that moment in the audio.
  • Tap the
    clock button
    on a session's transcript to show or hide timestamps
  • If the session's audio was saved, tap a time to play the conversation from just before that moment
  • Copying or downloading the transcript includes the times while they're shown
Where it works:
timestamps appear on sessions transcribed with
Nemotron
, live or imported. Sessions from Whisper or a cloud speech provider don't carry timing, so the clock button doesn't appear on them.
Tip:
turn on
Save Session Audio
when you start a session so you can jump to the audio later. On the web, timestamps show but aren't clickable.

new

macOS

iOS

Android

Web

Windows

Choose Your Reading Text Size (v3.12)

Make Hedy's reading surfaces as large or as small as you like.
  • A new slider in
    Settings → Privacy & System → Reading text size
    makes transcripts, notes, chat and highlights larger or smaller, from 80% to 120%
  • Speaker names in transcripts follow the same size
  • The setting is saved per device, so your phone and your computer can each have their own size

new

macOS

iOS

Android

Windows

xAI Speech Recognition (v3.12)

xAI joins Deepgram and OpenAI as a cloud speech provider you can use with your own API key.
  • Choose
    xAI
    in
    Settings → Speech & AI
    and add your xAI API key
  • Works for live sessions and for imported files
  • Your custom vocabulary is sent along as hints, so names and product terms come back spelled your way

improved

macOS

iOS

Android

Windows

Audio Player on the Transcript Tab (v3.12)

On Mac, Windows and tablets, the session's audio player now also sits on the
Transcript
tab, so you can listen while you read and check a passage against what was said. It appears whenever the session has saved audio.

improved

macOS

iOS

Android

Windows

More Reliable Cloud Speech Recognition (v3.12)

  • If the connection to Deepgram, OpenAI or xAI silently drops, Hedy notices within about a minute and switches to on-device speech recognition, so your transcript keeps going
  • Your custom vocabulary stays out of connection error messages and error reports

improved

Windows

Windows Updates Wait for Your Session (v3.12)

  • Starting with 3.12.0, an update waits while a session is in progress, being finished, or having its audio saved, then installs
  • The installer now lets Hedy close on its own rather than forcing it shut

improved

macOS

Windows

Microphone Mute Stays On (v3.12)

  • A muted microphone stays muted when you pause and resume a session, and when Hedy recovers audio capture
  • On Mac, it also stays muted when Hedy switches microphones automatically
  • System audio keeps being captured while your microphone is muted

improved

macOS

iOS

Android

Windows

Imports Keep Your Topic, Context and Language (v3.12)

  • Audio, video and transcript imports keep the topic and context you pick in the import dialog, including with Cloud Sync off
  • The language option follows your speech engine: with Nemotron English Only, it shows English

improved

macOS

iOS

Android

Windows

More Complete Transcripts After an Interruption (v3.12)

When Hedy recovers a session after an interruption, it keeps the full live transcript and speaker labels.

improved

macOS

iOS

Android

Web

Windows

Easier to Read (v3.11.1)

Summaries, notes and chat are easier to read.
  • Larger text
    for summaries, detailed notes and chat, with more line spacing in chat
  • One consistent reading size
    across transcripts, chat and session, topic and highlight cards
  • More room on phones:
    chat messages use the full width of the screen
  • The
    copy button
    under each chat answer is always visible again, on phones and tablets too
Load More
→