easwee
Build a voice agent using Soniox STT and TTS
Re:Ultima VII Revisited
on Wednesday November 19, 2025 @02:07AM (#45976660)
This brings back so much memories!
Re:Ask HN: Do you have an aversion to recent TLDs?
on Saturday November 01, 2025 @04:11AM (#45780010)
I don't - .com is boring.
Re:50 Things I Know
on Tuesday September 30, 2025 @03:18PM (#45429961)
You know a lot :)
Re:Baldur's Gate 3 Steam Deck – Native Version
on Wednesday September 24, 2025 @03:04AM (#45357219)
I finished the game on Ubuntu when it came out, so it should work fine since ever. You can check out https://www.protondb.com/app/1086940 for more info.
Re:Can't pay, won't pay: streaming services are driving viewers back t...
on Thursday August 14, 2025 @06:01PM (#44906206)
Driving back???
Re:Show HN: Compare Speech APIs Live (OpenAI, Google, Deepgram, Soniox...
on Tuesday July 15, 2025 @06:51AM (#44569833)
We built this to help devs compare real-time speech systems side by side.
You can speak into the mic or upload audio. The tool sends the same input to each provider and shows live outputs next to each other.
It’s fully open and reproducible. You can fork it on Github if you want to run your own local tests or plug in your own provider.
https://github.com/soniox/soniox-compare
Would love feedback or suggestions, especially from folks who have tested providers in production.
Show HN: Compare Speech APIs Live (OpenAI, Google, Deepgram, Soniox, etc.)
Re:It's still worth blogging in the age of AI
on Tuesday February 25, 2025 @07:34AM (#43171069)
I dont blog for others, I blog for myself and my joy.
Re:Collatz's Ant
on Monday December 23, 2024 @03:12AM (#42492739)
Lovely.
Re:Omnio: First AI model that can natively reason over audio
on Tuesday October 15, 2024 @11:11AM (#41849452)
I tried Eminem - Rap God with "transcribe word by word" prompt and it did quite good, I just had to set temperature to 0, else it can get creative.
Re:The Open Music Encyclopedia
on Monday September 30, 2024 @05:31PM (#41702397)
ytmp3 + musicbrainz picard + puddletag = <3
AI model that natively processes audio as the input modality to fully utilize all the information available in the audio signal.