Have something to say?
Feedback on Utter: meeting performance and app-triggered writing tools
I've been using Utter for meeting transcription with ElevenLabs Scribe v2 via BYOK and wanted to share some feedback along with a feature request. First, the positives: the meeting transcription feature is genuinely robust. Speaker renaming and segment reassignment work well and are better implemented than other tools I've used in this space. The ElevenLabs Scribe v2 integration via BYOK works cleanly and the diarization output is accurate. The overall meeting workflow is solid. One area that needs attention is performance in the meeting UI. When reviewing a recorded meeting, the interface becomes slow and choppy — scrolling through the transcript, switching between views, and editing speaker labels all feel sluggish. This is the main friction point in an otherwise well-built feature. Smoother performance in the meeting view would make a meaningful difference. On the feature request side: the Writing Tools system is well-conceived, but right now each tool requires its own dedicated shortcut. What I'd like to see is app and URL-triggered writing tools — the ability to group multiple apps or URLs under a single writing tool and have Utter automatically select the right tool based on where I am, all triggered by one global hotkey. For example, one hotkey fires a strip-trailing-periods tool when I'm in Messages, a formal cleanup tool when I'm in Outlook, and a default tool everywhere else, without me having to think about which shortcut to press. This pattern is already established in other dictation apps and it's genuinely transformative for daily workflow. If Utter can bring that to the Writing Tools system, it would be a very strong contender as a daily driver. I'm aware of the URL scheme for mode switching and have considered using Keyboard Maestro as a workaround, but a native implementation would be far more reliable and maintainable. Thanks for building a quality app. Happy to provide more detail on any of this.
Voice memos transcription issue
IMPORTANT NOTE: I have not been able to test this with voice memos recorded on the same device. My voice memos are recorded with the apple voice memo app on apple watch then I have attempted to transcribe using utter’s transcribe file function. Simply clicking and dragging the voice memo from finder or from the voice memo app results in this error code displaying. ios 18.7 macos sequoia (updated)
Press to record paste issue
On my Mac, I’ve configured the hold-to-record button to the Function key. If you hold down the key and don’t speak, releasing it pastes all the words you’ve configured in the app’s vocabulary section.
Ability to hide QWERTY keys from keyboard
The QWERTY keys aren’t as feature rich as the native keyboard. As such I toggle between Utter for dictation and Apple keyboard for everything else. Because Utter looks like the native keyboard it’s jarring when it doesn’t behave like it (swipe to type not supported). It would be great to be able to turn of QWERTY on Utter, so it’s just a dictation keyboard.
Google Docs issue?
I've been noticing that sometimes the app isn't working in Google Docs. The microphone icon will load, it'll look like it's capturing my dictation, but it won't paste anything. but if I move to a different application, like a notes app or the browser URL, it works again. It's an intermittent problem because sometimes it'll start working again, and then stop.
Auto-conversion of numbers to numeric values
It would be great if the app could natively convert spoken numbers to numeric values. So one thousand to 1,000 and so on.
Realtime transcription for local & online models
Realtime transcription for local & online models (BYOK & utter cloud). Should reduce latency by 50% to 0.5s
iOS Widgets, Shortcuts, Live Activity & Action Button
Live Activity for ongoing recordings Home Widget to see ongoing recording & start & stop recording Control Centre widget to start & stop recording Action Button Shortcut to start & stop recording with automatic copy to paste.
Notes: Edit Transcripts Using Markdown Editor
Add ability to edit transcripts using with simple markdown formatting.
AI Chat With Transcript
Chat with AI to ask questions about all transcripts, specific transcripts or all transcripts of a specific mode. RAG using local embedding models with query planner (SQLite FTS BM25 + Vector Search) and local reranker model
Smart Mode: Auto-Formatting Dictations
A transcription mode that automatically formats into dictations into common formats like email, notes, messages depending on context. Should reduce need for creating custom modes for common usecases and having to switch modes.
File Transcription Refresh: Batch File Transcription & Folder Watcher
Dedicated File Transcription Page Batch transcription for audio & video files with drag and drop interface. Folder watcher for automatic transcription of files added to selected folders.
System Audio Recording For Meetings
Adding option to record system audio in addition to device input option for meetings.
Youtube Video Summarizer
Summarize any Youtube video, PDF/DOCX/PPT file, Websites, Podcasts (RSS Feed) with a link and AI chat to ask questions about the content.