So I have turned to a complex and highly unreliable software stack that provides both voice-to-text, and clumsy but limited control of Microsoft Windows, Chrome, etc. This includes Dragon Voice-to-text, Voice Computer, and Talon, plus a browser extension and heavy customization.
Users of Dragon will acknowledge that:\na) The software is a creaky dumpster fire built on archaic code\nb) There is no viable alternative on the market
My question is: *how is it that no one has built something better?* The market is huge, and the Natural Language Processing of "OK Google" and Siri are quite refined at this point.
References:
Dragon: https://www.nuance.com/dragon.html
Voice Computer: https://voicecomputer.com/
Talon: https://talonvoice.com/"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Why is there no high quality method for voice control of a PC?"}},"_tags":["story","author_4midori","story_30117383","ask_hn"],"author":"4midori","children":[30117535,30117600,30117602,30117661,30117718,30117727,30117767,30117783,30117786,30117789,30117807,30117824,30117864,30117972,30117995,30118030,30118133,30118138,30118142,30118156,30118201,30118224,30118237,30118340,30118348,30118562,30119014,30119659,30119766,30120282,30120338,30120375,30120498,30120607,30120650,30121175,30121564,30121746,30122669,30122706,30123191,30123648,30124638,30124642,30124647,30124730,30125514,30132029,30139942],"created_at":"2022-01-28T17:05:48Z","created_at_i":1643389548,"num_comments":119,"objectID":"30117383","points":90,"story_id":30117383,"story_text":"Like many people who have spent decades behind a keyboard, RSI (Repetitive Stress Injury) prevents me from writing code and doing graphic design through the usual keyboard and mouse inputs.
So I have turned to a complex and highly unreliable software stack that provides both voice-to-text, and clumsy but limited control of Microsoft Windows, Chrome, etc. This includes Dragon Voice-to-text, Voice Computer, and Talon, plus a browser extension and heavy customization.
Users of Dragon will acknowledge that:\na) The software is a creaky dumpster fire built on archaic code\nb) There is no viable alternative on the market
My question is: *how is it that no one has built something better?* The market is huge, and the Natural Language Processing of "OK Google" and Siri are quite refined at this point.
References:
Dragon: https://www.nuance.com/dragon.html
Voice Computer: https://voicecomputer.com/
Talon: https://talonvoice.com/","title":"Ask HN: Why is there no high quality method for voice control of a PC?","updated_at":"2024-09-20T10:21:34Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Tsarp"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"About a year ago, my RSI had gotten to the point where I needed to take days off. I went down the rabbit hole of trackball mice, split keyboards, and dictation. Talon voice coding seemed like the dream solution I was looking for, but I just couldn\u2019t get it right (here\u2019s a link to the video I was inspired by: https://www.youtube.com/watch?v=0ZZb12Qp6-0)). Around the same time, I increasingly adapted to new ways of life- tabbing, code suggests, completions, chat-based coding (I know, I know), and brainstorming with tools like Copilot, Codeium, Claude, and then Cursor.
Mac's built-in dictation was my first attempt at a solution. It handled basic emails/messages fine but completely fell apart with technical terms. I couldn\u2019t build a workflow around it. \u201cmalloc,\u201d "memcpy," \u201cAxum,\u201d \u201cTauri\u201d\u2026 these always had to be edited, and my wrists would still protest.
I had been playing around with Whisper ever since it first came out (my work with MEL spectrograms and ECGs led me deep into OpenAI's Whisper). Once the whisper.cpp project got metal support, it worked beautifully on Apple Silicon Macs.
What I ended up building
- Local dictation tool
- Works system-wide on Mac (any text input)
- Custom dictionaries for technical terms/jargon
- No cloud dependencies, no tracking, no signup, no subscriptions (one-time purchase)
Technical details
- Running Whisper Large Turbo locally (or quantized Whisper small/medium)
- Sits in the background, requests accessibility permissions, and can paste text anywhere
- Voice activity detection is experimental; haven't nailed the parameters yet. This will help with long recordings/live transcriptions
- Sandboxed app, available through the Mac App Store
My usage stats (just personal benchmarks)
- Transcribed around 30k words over 2 weeks
- Averaging 150 WPM vs my usual 55-60 WPM (significantly lower when I need to think\u2014it\u2019s not like one of those typing tests)
- I\u2019ve spoken for about 4 hours to it. So roughly 6-8 hours saved
- Some occasional mistakes; there's a find and replace tool. I add random non-dictionary words for large prompts
https://carelesswhisper.app/"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Solution to RSI and poor Mac dictation, accurately transcribes \"memcpy\""},"url":{"matchLevel":"none","matchedWords":[],"value":"https://carelesswhisper.app/"}},"_tags":["story","author_Tsarp","story_42460676","show_hn"],"author":"Tsarp","children":[42463421,42463524,42464026,42514512],"created_at":"2024-12-19T12:09:05Z","created_at_i":1734610145,"num_comments":4,"objectID":"42460676","points":8,"story_id":42460676,"story_text":"About a year ago, my RSI had gotten to the point where I needed to take days off. I went down the rabbit hole of trackball mice, split keyboards, and dictation. Talon voice coding seemed like the dream solution I was looking for, but I just couldn\u2019t get it right (here\u2019s a link to the video I was inspired by: https://www.youtube.com/watch?v=0ZZb12Qp6-0)). Around the same time, I increasingly adapted to new ways of life- tabbing, code suggests, completions, chat-based coding (I know, I know), and brainstorming with tools like Copilot, Codeium, Claude, and then Cursor.
Mac's built-in dictation was my first attempt at a solution. It handled basic emails/messages fine but completely fell apart with technical terms. I couldn\u2019t build a workflow around it. \u201cmalloc,\u201d "memcpy," \u201cAxum,\u201d \u201cTauri\u201d\u2026 these always had to be edited, and my wrists would still protest.
I had been playing around with Whisper ever since it first came out (my work with MEL spectrograms and ECGs led me deep into OpenAI's Whisper). Once the whisper.cpp project got metal support, it worked beautifully on Apple Silicon Macs.
What I ended up building
- Local dictation tool
- Works system-wide on Mac (any text input)
- Custom dictionaries for technical terms/jargon
- No cloud dependencies, no tracking, no signup, no subscriptions (one-time purchase)
Technical details
- Running Whisper Large Turbo locally (or quantized Whisper small/medium)
- Sits in the background, requests accessibility permissions, and can paste text anywhere
- Voice activity detection is experimental; haven't nailed the parameters yet. This will help with long recordings/live transcriptions
- Sandboxed app, available through the Mac App Store
My usage stats (just personal benchmarks)
- Transcribed around 30k words over 2 weeks
- Averaging 150 WPM vs my usual 55-60 WPM (significantly lower when I need to think\u2014it\u2019s not like one of those typing tests)
- I\u2019ve spoken for about 4 hours to it. So roughly 6-8 hours saved
- Some occasional mistakes; there's a find and replace tool. I add random non-dictionary words for large prompts
https://carelesswhisper.app/","title":"Show HN: Solution to RSI and poor Mac dictation, accurately transcribes \"memcpy\"","updated_at":"2025-08-20T08:33:58Z","url":"https://carelesswhisper.app/"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"talonbragg"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"This is an html framework. You can access the github repo here: https://github.com/talonbragg/Custom-Markup. You can also visit the website here: https://custommarkup.ml"},"title":{"matchLevel":"none","matchedWords":[],"value":"Custom Markup"}},"_tags":["story","author_talonbragg","story_13346744","ask_hn"],"author":"talonbragg","children":[13346851,13349266,13351829],"created_at":"2017-01-07T21:26:47Z","created_at_i":1483824407,"num_comments":3,"objectID":"13346744","points":3,"story_id":13346744,"story_text":"This is an html framework. You can access the github repo here: https://github.com/talonbragg/Custom-Markup. You can also visit the website here: https://custommarkup.ml","title":"Custom Markup","updated_at":"2024-09-20T00:13:23Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"talonbragg"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"This html framework allows you to style your code without having to use your own css. The link to the site is: https://custommarkup.ml. The link to the github repo is: https://github.com/talonbragg/Custom-Markup."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Custom Markup, a Simple HTML Framework"}},"_tags":["story","author_talonbragg","story_13365744","show_hn"],"author":"talonbragg","created_at":"2017-01-10T15:56:16Z","created_at_i":1484063776,"num_comments":0,"objectID":"13365744","points":3,"story_id":13365744,"story_text":"This html framework allows you to style your code without having to use your own css. The link to the site is: https://custommarkup.ml. The link to the github repo is: https://github.com/talonbragg/Custom-Markup.","title":"Show HN: Custom Markup, a Simple HTML Framework","updated_at":"2024-09-20T00:15:23Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"shukantpal"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"Pretzel is an experiment in live AI agents for music. It's a web synchronized music sequencer that is controlled by an AI agent. Everyone chats with the same AI agent and hears the same music at any time. I'm no audio engineer so it's a pretty basic tune, and turns out audio is a pretty complex 3d function.
I went to Google's IO hackathon yesterday without any idea of what I was going to work on. I thought of building a live DJ agent that people can tell what they feel and it would update the music live. This is like 50% of the way that vision.
I've been building a Rust agent harness called [Talon](https://github.com/impalasys/talon) that is easy to self host on a micro VM (~100mb RAM). Pretzel connects everyone to the same session in a Talon hosted DJ agent.
I'm curious what y'all think."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Live AI music sequencing agent"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://pretzel.shukant.com/?nickname=Anonymous&role=stage"}},"_tags":["story","author_shukantpal","story_48263249","show_hn"],"author":"shukantpal","created_at":"2026-05-25T03:53:09Z","created_at_i":1779681189,"num_comments":0,"objectID":"48263249","points":2,"story_id":48263249,"story_text":"Pretzel is an experiment in live AI agents for music. It's a web synchronized music sequencer that is controlled by an AI agent. Everyone chats with the same AI agent and hears the same music at any time. I'm no audio engineer so it's a pretty basic tune, and turns out audio is a pretty complex 3d function.
I went to Google's IO hackathon yesterday without any idea of what I was going to work on. I thought of building a live DJ agent that people can tell what they feel and it would update the music live. This is like 50% of the way that vision.
I've been building a Rust agent harness called [Talon](https://github.com/impalasys/talon) that is easy to self host on a micro VM (~100mb RAM). Pretzel connects everyone to the same session in a Talon hosted DJ agent.
I'm curious what y'all think.","title":"Show HN: Live AI music sequencing agent","updated_at":"2026-05-25T08:40:04Z","url":"https://pretzel.shukant.com/?nickname=Anonymous&role=stage"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pandaplaysall"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"Sernade is discontinued, Dragon is discontinued leaving Mac dictation and Talon.
What\u2019s the best approach to coding voice only and controlling the Mac currently?
Also what about IDEs? VSCODE has cursorless, is it good? IntelliJ only has a talon plug-in that has bad reviews.
So any coding plugins for IntelliJ and Neovim as well?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Voice Coding/Control on Mac"}},"_tags":["story","author_pandaplaysall","story_38976335","ask_hn"],"author":"pandaplaysall","created_at":"2024-01-13T01:08:07Z","created_at_i":1705108087,"num_comments":0,"objectID":"38976335","points":2,"story_id":38976335,"story_text":"Sernade is discontinued, Dragon is discontinued leaving Mac dictation and Talon.
What\u2019s the best approach to coding voice only and controlling the Mac currently?
Also what about IDEs? VSCODE has cursorless, is it good? IntelliJ only has a talon plug-in that has bad reviews.
So any coding plugins for IntelliJ and Neovim as well?","title":"Voice Coding/Control on Mac","updated_at":"2024-09-20T16:07:59Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"speckx"},"title":{"matchLevel":"none","matchedWords":[],"value":"Stratolaunch Successfully Completes Reusable Hypersonic Flight and Recovery"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"https://www.stratolaunch.com/news/stratolaunch-successfully-completes-reusable-hypersonic-flight-and-recovery-with-talon-a2-vehicle/"}},"_tags":["story","author_speckx","story_43914852"],"author":"speckx","children":[43943008,43943075,43944623,43944972],"created_at":"2025-05-07T12:35:23Z","created_at_i":1746621323,"num_comments":13,"objectID":"43914852","points":63,"story_id":43914852,"title":"Stratolaunch Successfully Completes Reusable Hypersonic Flight and Recovery","updated_at":"2025-05-12T10:04:15Z","url":"https://www.stratolaunch.com/news/stratolaunch-successfully-completes-reusable-hypersonic-flight-and-recovery-with-talon-a2-vehicle/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"belter"},"title":{"matchLevel":"none","matchedWords":[],"value":"Stratolaunch reveals its first hypersonic design for high-altitude flights"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"https://www.space.com/stratolaunch-talon-a-vehicle-debut"}},"_tags":["story","author_belter","story_31569934"],"author":"belter","children":[31571521,31574915,31575654,31575728,31577317],"created_at":"2022-05-31T14:32:43Z","created_at_i":1654007563,"num_comments":17,"objectID":"31569934","points":40,"story_id":31569934,"title":"Stratolaunch reveals its first hypersonic design for high-altitude flights","updated_at":"2024-09-20T11:17:26Z","url":"https://www.space.com/stratolaunch-talon-a-vehicle-debut"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"The bounty hunters of Fallout 76 will kill anyone you want, for a price"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"https://www.polygon.com/2019/7/12/20690651/fallout-76-talon-company-appalachia-bounty-bethesda-group"}},"_tags":["story","author_tosh","story_20421885"],"author":"tosh","children":[20422235,20422290,20424097],"created_at":"2019-07-12T16:24:37Z","created_at_i":1562948677,"num_comments":6,"objectID":"20421885","points":27,"story_id":20421885,"title":"The bounty hunters of Fallout 76 will kill anyone you want, for a price","updated_at":"2024-09-20T04:33:57Z","url":"https://www.polygon.com/2019/7/12/20690651/fallout-76-talon-company-appalachia-bounty-bethesda-group"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"talonbragg"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: GateKeeperCLI: environment variables done right"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"https://github.com/talonbragg/gatekeepercli"}},"_tags":["story","author_talonbragg","story_22677443","show_hn"],"author":"talonbragg","children":[22705205],"created_at":"2020-03-24T18:35:29Z","created_at_i":1585074929,"num_comments":2,"objectID":"22677443","points":8,"story_id":22677443,"title":"Show HN: GateKeeperCLI: environment variables done right","updated_at":"2024-09-20T05:55:16Z","url":"https://github.com/talonbragg/gatekeepercli"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"talonbragg"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Gatekeepercli: environment variables done right"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["talon"],"value":"https://github.com/talonbragg/gatekeepercli"}},"_tags":["story","author_talonbragg","story_22722784","show_hn"],"author":"talonbragg","created_at":"2020-03-29T21:11:53Z","created_at_i":1585516313,"num_comments":0,"objectID":"22722784","points":5,"story_id":22722784,"title":"Show HN: Gatekeepercli: environment variables done right","updated_at":"2024-09-20T05:50:50Z","url":"https://github.com/talonbragg/gatekeepercli"}],"hitsPerPage":50,"nbHits":22559,"nbPages":20,"page":0,"params":"query=Talon&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":17,"processingTimingsMS":{"_request":{"roundTrip":24},"afterFetch":{"format":{"total":1},"merge":{"mergeLoop":{"prepareNextHit":3,"total":3},"total":3},"total":3},"fetch":{"query":3,"scanning":9,"total":13},"total":17},"query":"Talon","serverTimeMS":19}