Completed
Please submit your application to Apple for Notarization
I'm dying to use your application, but granting access to control my computer and microphone without being able to inspect/run source code and/or having Apple Notarize the App is just not tractable. If you're not going to share the source code, then please submit the code to Apple so that I can confidently install the application without worrying that it's full of malware.

Colin Fish 4 days ago
Completed
Please submit your application to Apple for Notarization
I'm dying to use your application, but granting access to control my computer and microphone without being able to inspect/run source code and/or having Apple Notarize the App is just not tractable. If you're not going to share the source code, then please submit the code to Apple so that I can confidently install the application without worrying that it's full of malware.

Colin Fish 4 days ago
In Progress
Automatic translation to target language for supported cloud models
Cloud models like Deepgram, ElevenLabs, AssemblyAI, Soniox support translation from a source language to a target language. For example, if I’m speaking in french and I want the transcribed text to be automatically translated to English before pasting in the input box.

vowen_user 4 days ago
In Progress
Automatic translation to target language for supported cloud models
Cloud models like Deepgram, ElevenLabs, AssemblyAI, Soniox support translation from a source language to a target language. For example, if I’m speaking in french and I want the transcribed text to be automatically translated to English before pasting in the input box.

vowen_user 4 days ago
In Progress
Make recording indicator work on multi screen setup
When there are multiple screens setup, make the recording indicator show up specifically in the screen where the cursor is

Vowen 4 days ago
In Progress
Make recording indicator work on multi screen setup
When there are multiple screens setup, make the recording indicator show up specifically in the screen where the cursor is

Vowen 4 days ago
In Progress
Request for Whisper 3 Large Support in Romanian
Please kindly add support also for Whisper 3 large, for groq, because it works much better in Romanian. Also, another thing: if you speak a little faster, the words get mixed up. Also, AI doesn't understands very well. Make it so that you can use AI to detail a prompt of 100-200 words, because now it gives you everything with JSON formatting. Thank you very much.

Radu 4 days ago
In Progress
Request for Whisper 3 Large Support in Romanian
Please kindly add support also for Whisper 3 large, for groq, because it works much better in Romanian. Also, another thing: if you speak a little faster, the words get mixed up. Also, AI doesn't understands very well. Make it so that you can use AI to detail a prompt of 100-200 words, because now it gives you everything with JSON formatting. Thank you very much.

Radu 4 days ago
Completed
Final words are not captured
Sometimes when ending the transcription right after the phrase ends, it'll miss the last word or a couple of words that were said. I've tried to get used to waiting a second before letting go of the record keyboard shortcut, but a lot of the time it's just instinctive to let go as soon as you're done.

Adrian Pauly 5 days ago
Completed
Final words are not captured
Sometimes when ending the transcription right after the phrase ends, it'll miss the last word or a couple of words that were said. I've tried to get used to waiting a second before letting go of the record keyboard shortcut, but a lot of the time it's just instinctive to let go as soon as you're done.

Adrian Pauly 5 days ago
Completed
Hands free mode(?) / Accessibility issue with sites like Google Docs
Hi! i've been trying to use the software as an alternative to Wispr. I noticed when I'm using hands free mode, haven't tested it out with the other modes because of physical disability - when I tap the hotkey to have it start transcribing it will stop registering me as on a Google doc for example. The blinking typing indicator will freeze essentially. Like, I will have to re click on the google doc to have it be aware to actually send the transcribed material onto it. If I don't, nothing will be typed and I'll have to redo everything. I'm using this with foot pedals as an accessibility aid because I can't type well due to issues with my wrists so having to click to coordinate where my words are sent is a bit of an issue!

Throwaway5325 7 days ago
Completed
Hands free mode(?) / Accessibility issue with sites like Google Docs
Hi! i've been trying to use the software as an alternative to Wispr. I noticed when I'm using hands free mode, haven't tested it out with the other modes because of physical disability - when I tap the hotkey to have it start transcribing it will stop registering me as on a Google doc for example. The blinking typing indicator will freeze essentially. Like, I will have to re click on the google doc to have it be aware to actually send the transcribed material onto it. If I don't, nothing will be typed and I'll have to redo everything. I'm using this with foot pedals as an accessibility aid because I can't type well due to issues with my wrists so having to click to coordinate where my words are sent is a bit of an issue!

Throwaway5325 7 days ago
Completed
Empty transcription = .
Another very minor one: if I use the push-to-talk mode and then let it go without saying anything, it will output a period and a line break. This is slightly annoying because sometimes I'll let go of the keys because I changed my mind and I didn't want to input anything yet. And then it will input a period which I have to go and delete. . . . It seems that here it's adding a period and a new line. Doing it now → . .

Adrian Pauly 9 days ago
Completed
Empty transcription = .
Another very minor one: if I use the push-to-talk mode and then let it go without saying anything, it will output a period and a line break. This is slightly annoying because sometimes I'll let go of the keys because I changed my mind and I didn't want to input anything yet. And then it will input a period which I have to go and delete. . . . It seems that here it's adding a period and a new line. Doing it now → . .

Adrian Pauly 9 days ago
Completed
"Thank you."
This one is not a huge deal but it’s weird and slightly funny. Occasionally the transcript will just add a “Thank you.” at the end of of whatever I just said. Sometimes if I activate the transcription keyboard shortcut in hold-to-talk mode and then let go without saying anything, it will output just the “Thank you.” Maybe you guys just wanted it to be super polite? If that’s the case you could add a locale setting and if the user selects Canada then output a random “Sorry!” from time to time

Adrian Pauly 9 days ago
Completed
"Thank you."
This one is not a huge deal but it’s weird and slightly funny. Occasionally the transcript will just add a “Thank you.” at the end of of whatever I just said. Sometimes if I activate the transcription keyboard shortcut in hold-to-talk mode and then let go without saying anything, it will output just the “Thank you.” Maybe you guys just wanted it to be super polite? If that’s the case you could add a locale setting and if the user selects Canada then output a random “Sorry!” from time to time

Adrian Pauly 9 days ago
Completed
No way to cancel hands-free mode with keyboard
When using hands-free mode, if I change my mind, there's no way to cancel the transcription. I am not 100% sure of this, but I think that Wispr Flow allowed canceling out of the transcription with Esc or something similar.

Adrian Pauly 13 days ago
Completed
No way to cancel hands-free mode with keyboard
When using hands-free mode, if I change my mind, there's no way to cancel the transcription. I am not 100% sure of this, but I think that Wispr Flow allowed canceling out of the transcription with Esc or something similar.

Adrian Pauly 13 days ago
Completed
Pause/stop buttons don't work in Windows
The stop and pause buttons that appear in the overlay indicator when dictating in hands-off mode don't do anything.

Adrian Pauly 13 days ago
Completed
Pause/stop buttons don't work in Windows
The stop and pause buttons that appear in the overlay indicator when dictating in hands-off mode don't do anything.

Adrian Pauly 13 days ago
Completed
User's dictionary should be A-z sorted
…and deduplicated. Mine does have about 40-50 entries (bun/ts/js/.env and all those IT silly slang) and some of them already duplicated. Pretty please?

Maksym Brashchenko 14 days ago
Completed
User's dictionary should be A-z sorted
…and deduplicated. Mine does have about 40-50 entries (bun/ts/js/.env and all those IT silly slang) and some of them already duplicated. Pretty please?

Maksym Brashchenko 14 days ago
Completed
local Voxtral engine?
https://simonwillison.net/2026/Feb/4/voxtral-2/ https://mistral.ai/news/voxtral-transcribe-2 https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602

Phil Wolff 20 days ago
Completed
local Voxtral engine?
https://simonwillison.net/2026/Feb/4/voxtral-2/ https://mistral.ai/news/voxtral-transcribe-2 https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602

Phil Wolff 20 days ago
Completed
Start a new paragraph with custom instructions
Ciao guys, love Vowen, coming from Wispr Flow. The biggest missing feature, for an e-mail writer like me, is for sure the skill of creating a new paragraph when it recognizes a new topic. I tried to force it with Custom Instruction (paste below it), but I’m still forced to manually refine it before sending. Am I missing something maybe, or could this be considered as a new feature? For context: I’m using Whisper Large v3 Turbo by Groq + Gemini 2.5 flash lite via Openrouter. Thanks! Custom Instruction given: { "system_instruction": { "role_definition": { "role": "Structural Text Architect & UX Content Editor", "objective": "Transform raw, noisy voice-to-text transcriptions into professionally formatted Italian text with aggressive paragraph segmentation and zero leading whitespace." }, "technical_constraints": { "no_leading_spaces": "CRITICAL: The output must start exactly at character index 0. Absolutely no spaces, tabs, or newlines before the first word.", "aggressive_paragraphing": "Force segment the text into short, readable paragraphs using double newlines (\\n\\n). Trigger a new paragraph at every topic shift or transition word (e.g., 'Inoltre', 'Quindi', 'Per quanto riguarda').", "signature_rule": "In email closings, the name must be on a new line and MUST NOT be followed by a period or any punctuation.", "clean_output_only": "Output only the processed text. Do not include conversational filler, meta-talk, or introductions." }, "operational_workflow": { "step_1": "De-noising: Remove Italian vocal fillers (ehm, allora, cioè, diciamo) and speech self-corrections.", "step_2": "Syntax Repair: Fix Italian grammar, verb agreements, and punctuation within sentences.", "step_3": "Structural Splitting: Identify semantic breaks and apply double newlines (\\n\\n) to ensure visual separation.", "step_4": "Signature Sanitization: Ensure the final line (the name) is punctuation-free.", "step_5": "Final Trimming: Strip all leading/trailing whitespace from the entire payload." }, "few_shot_examples": [ { "input": "buongiorno allora per il progetto va bene tutto però dobbiamo aggiungere quella clausola inoltre mi serve che mi mandi i documenti grazie claudio.", "output": "Buongiorno,\n\nper il progetto va bene tutto, però dobbiamo aggiungere quella clausola.\n\nInoltre, mi serve che mi invii i documenti.\n\nGrazie\nClaudio" } ], "output_configuration": { "encoding": "UTF-8", "newline_character": "\\n\\n", "start_index": 0 } } }

Claudio Grimoldi 22 days ago
Completed
Start a new paragraph with custom instructions
Ciao guys, love Vowen, coming from Wispr Flow. The biggest missing feature, for an e-mail writer like me, is for sure the skill of creating a new paragraph when it recognizes a new topic. I tried to force it with Custom Instruction (paste below it), but I’m still forced to manually refine it before sending. Am I missing something maybe, or could this be considered as a new feature? For context: I’m using Whisper Large v3 Turbo by Groq + Gemini 2.5 flash lite via Openrouter. Thanks! Custom Instruction given: { "system_instruction": { "role_definition": { "role": "Structural Text Architect & UX Content Editor", "objective": "Transform raw, noisy voice-to-text transcriptions into professionally formatted Italian text with aggressive paragraph segmentation and zero leading whitespace." }, "technical_constraints": { "no_leading_spaces": "CRITICAL: The output must start exactly at character index 0. Absolutely no spaces, tabs, or newlines before the first word.", "aggressive_paragraphing": "Force segment the text into short, readable paragraphs using double newlines (\\n\\n). Trigger a new paragraph at every topic shift or transition word (e.g., 'Inoltre', 'Quindi', 'Per quanto riguarda').", "signature_rule": "In email closings, the name must be on a new line and MUST NOT be followed by a period or any punctuation.", "clean_output_only": "Output only the processed text. Do not include conversational filler, meta-talk, or introductions." }, "operational_workflow": { "step_1": "De-noising: Remove Italian vocal fillers (ehm, allora, cioè, diciamo) and speech self-corrections.", "step_2": "Syntax Repair: Fix Italian grammar, verb agreements, and punctuation within sentences.", "step_3": "Structural Splitting: Identify semantic breaks and apply double newlines (\\n\\n) to ensure visual separation.", "step_4": "Signature Sanitization: Ensure the final line (the name) is punctuation-free.", "step_5": "Final Trimming: Strip all leading/trailing whitespace from the entire payload." }, "few_shot_examples": [ { "input": "buongiorno allora per il progetto va bene tutto però dobbiamo aggiungere quella clausola inoltre mi serve che mi mandi i documenti grazie claudio.", "output": "Buongiorno,\n\nper il progetto va bene tutto, però dobbiamo aggiungere quella clausola.\n\nInoltre, mi serve che mi invii i documenti.\n\nGrazie\nClaudio" } ], "output_configuration": { "encoding": "UTF-8", "newline_character": "\\n\\n", "start_index": 0 } } }

Claudio Grimoldi 22 days ago
Rejected
Make it Open Source
Hey. Thanks for the awesome tool. Not exactly a feature request. And I believe you have your reasons for not open sourcing it. But nevertheless I wanted to share a perspective you may not have considered. There are a ton of other similar solutions out there of various degrees of polish. And the heart of all of them (including Vowen) is those open-weight models anyway. So the fundamental technology you and others use is not a trade secret. While commitment to do the processing locally and/or with the user-configured providers, and not send user data to the developers sounds great, it's not really easily verifiable. It basically makes Vowen impossible to get a permission to use Vowen in an enterprise environment where information security standards are high. For closed source solutions the matter of trust is often solved through certifications (SOC2, ISO etc). However this is clearly not a viable solution for a side project worked by a team of 2 (both due to overhead to manage processes and cost of such certifications). This effectively narrows down the possible user base. Now,if you're looking to monetize your app in the future, it may be worth to learn from many AI-powered open source tools out there (OpenCode, Kilo Code etc.). They often have the end user app open source, but provide paid a control plane for enterprises to centrally manage their users and their licenses instead of having to issue and rotate individual tokens to each employee. So, I hope you'll at least give this idea another thought. Thanks again for building Vowen. Alex

Alexandr Zolotukhin 24 days ago
Rejected
Make it Open Source
Hey. Thanks for the awesome tool. Not exactly a feature request. And I believe you have your reasons for not open sourcing it. But nevertheless I wanted to share a perspective you may not have considered. There are a ton of other similar solutions out there of various degrees of polish. And the heart of all of them (including Vowen) is those open-weight models anyway. So the fundamental technology you and others use is not a trade secret. While commitment to do the processing locally and/or with the user-configured providers, and not send user data to the developers sounds great, it's not really easily verifiable. It basically makes Vowen impossible to get a permission to use Vowen in an enterprise environment where information security standards are high. For closed source solutions the matter of trust is often solved through certifications (SOC2, ISO etc). However this is clearly not a viable solution for a side project worked by a team of 2 (both due to overhead to manage processes and cost of such certifications). This effectively narrows down the possible user base. Now,if you're looking to monetize your app in the future, it may be worth to learn from many AI-powered open source tools out there (OpenCode, Kilo Code etc.). They often have the end user app open source, but provide paid a control plane for enterprises to centrally manage their users and their licenses instead of having to issue and rotate individual tokens to each employee. So, I hope you'll at least give this idea another thought. Thanks again for building Vowen. Alex

Alexandr Zolotukhin 24 days ago
Completed
Unexpected behaviors in keyboard shortcut config in Windows
A couple of unexpected behaviors when trying to configure my shortucts in the Windows app: The keys show up as Mac keys (command, option, etc). I couldn’t find a setting to change them but then again, this is a Windows app, they should be Windows keys by default, no? I set up the transcription hotkey shortcut (ctrl+Win). Then I wanted to set Ctrl+Win+Space as my hands-free shortcut, emulating what I had in Wispr Flow. But when I press the combination to try to set the hands-free, since it´s already assigned to the transcription shortcut, it will activate the transcription and when I let go, it adds (Command + V) to the shortcut. After some fiddling around I discovered that I can press the keys one by one and build the shortcut that way, which worked, but it wasn’t my first thought to do it like that and I´m pretty sure for most apps you just press the combination of keys as you would when activating the shortcut. Even after setting the shorcuts the way I wanted, they don´t work as expected or as they do in Wispr Flow which I found super useful, this would be a nice addition! 1. You press Ctrl+Win to activate transcription 2. While those are held down, you can hit spacebar to “lock” it into hands-free mode 3. If you’re holding down Ctrl+Win after some time (it’s probably like 10 seconds or so) it will show a small message above the audio waveform indicator suggesting that you can press spacebar to lock it into hands-free mode.

Adrian Pauly 25 days ago
Completed
Unexpected behaviors in keyboard shortcut config in Windows
A couple of unexpected behaviors when trying to configure my shortucts in the Windows app: The keys show up as Mac keys (command, option, etc). I couldn’t find a setting to change them but then again, this is a Windows app, they should be Windows keys by default, no? I set up the transcription hotkey shortcut (ctrl+Win). Then I wanted to set Ctrl+Win+Space as my hands-free shortcut, emulating what I had in Wispr Flow. But when I press the combination to try to set the hands-free, since it´s already assigned to the transcription shortcut, it will activate the transcription and when I let go, it adds (Command + V) to the shortcut. After some fiddling around I discovered that I can press the keys one by one and build the shortcut that way, which worked, but it wasn’t my first thought to do it like that and I´m pretty sure for most apps you just press the combination of keys as you would when activating the shortcut. Even after setting the shorcuts the way I wanted, they don´t work as expected or as they do in Wispr Flow which I found super useful, this would be a nice addition! 1. You press Ctrl+Win to activate transcription 2. While those are held down, you can hit spacebar to “lock” it into hands-free mode 3. If you’re holding down Ctrl+Win after some time (it’s probably like 10 seconds or so) it will show a small message above the audio waveform indicator suggesting that you can press spacebar to lock it into hands-free mode.

Adrian Pauly 25 days ago
Completed
Auto-stop dictation when typing or clicking outside the text box
Please add an option for dictation to stop automatically when I start typing or click outside the active text box — like Windows Speech-to-Text. This makes it easier to move between multiple text boxes or screens without pressing the shortcut to stop in the middle each time.

Saad Elg 29 days ago
Completed
Auto-stop dictation when typing or clicking outside the text box
Please add an option for dictation to stop automatically when I start typing or click outside the active text box — like Windows Speech-to-Text. This makes it easier to move between multiple text boxes or screens without pressing the shortcut to stop in the middle each time.

Saad Elg 29 days ago