logoalt Hacker News

Show HN: Ghost Pepper – Local hold-to-talk speech-to-text for macOS

462 points • by MattHart88 • last Monday at 7:50 PM • 195 comments • view on HN

I built this because I wanted to see how far I could get with a voice-to-text app that used 100% local models so no data left my computer. I've been using a ton for coding and emails. Experimenting with using it as a voice interface for my other agents too. 100% open-source MIT license, would love feedback, PRs, and ideas on where to take it.

Comments

ssz7820 • last Tuesday at 6:04 PM

[dead]

cisco801 • yesterday at 4:30 AM

[dead]

edoardobambini- • last Tuesday at 11:14 AM

[dead]

jiusanzhou • last Tuesday at 7:02 AM

[dead]

aesopturtle • last Tuesday at 2:09 AM

[dead]

Sukhbat • last Tuesday at 5:45 AM

[flagged]

rpdaiml • last Tuesday at 1:17 PM

[dead]

atlasagentsuite • last Monday at 11:53 PM

[flagged]

zhichuanxun • last Tuesday at 2:54 PM

[dead]

joshuahart • last Tuesday at 8:49 AM

[dead]

zephyrwhimsy • last Tuesday at 1:24 AM

[flagged]

bambushu • last Tuesday at 7:25 AM

nice to see this running fully local. what model size are you shipping as default, and what's the cold-start time on Apple Silicon? I've been using Whisper locally for meeting transcription and the biggest friction point is always endpoint detection - knowing when you've stopped talking vs pausing to think. curious how you handle that with hold-to-talk.

zephyrwhimsy • last Tuesday at 9:09 AM

[flagged]

Ecko123 • last Monday at 10:07 PM

[dead]

eddie-wang • last Tuesday at 6:18 AM

[dead]

lostathome • last Monday at 8:59 PM

[dead]

philbitt • last Tuesday at 2:17 PM

[dead]

romeroej • last Monday at 9:51 PM

always mac. when windows? why can you just make things multios

➕ show 4 replies