Replying to @charliermarsh
Hey Charlie, I got the $100 grant about a month ago, but my account never got upgraded to Pro 200. I can't wait to try the new ultrafast mode. Any chance you could take a look? Happy to DM my account email.
1
1
19
Hebrew support runs on ReNikud, a small model that adds the missing vowels so the synth knows how to pronounce each word.
PR for @jamesonhodge: github.com/jamesondh/chimera…
ReNikud: github.com/renikud/renikud
3
92
CHIMERA sings in Hebrew now!! 🇮🇱
Hebrew is written without vowels so getting this to work was fun :)
turn your sound on, details below 🧵
cc @bcherny @lydiahallie
would love to see something like this built into Claude Code ;)
2
24
Pretty much instantly! Thankfully macOS has a built-in echo cancellation feature in it's recording API.
1
1
16
Open source, MIT. macOS on Apple silicon, first run downloads ~700MB of models.
github.com/thewh1teagle/hand…
Try it and tell me what breaks 🐛
2
2
100
I built a voice mode for Claude Code.
Like a phone call: speak, listen, interrupt. No push-to-talk.
STT and TTS run locally on your Mac (Parakeet + Phonon).
One command:
/plugin install handsfree --marketplace thewh1teagle/handsfree
@bchern @GradiumAI 👇
Replying to @BrianRoemmele
The US doesn't seem to understand that the genie is already out of the bottle
1
31
Replying to @Neuralwarlock
Great read! Could JEPA architecture help bridge the gap between LLMs and speech?
1
133
Replying to @YehoyadaDev
חושב על זה הרבה גם, אני מאמין שהתמזל מזלנו שאנחנו (סטטיסטית) עם המוחות החזקים בעולם אז איכשהו זה יסתדר בישראל
2
109

