News for hackers. Stuff that matters.

A comment on DeepSeek Introduces Vision. Parent comment

Re:DeepSeek Introduces Vision by dghlsakjg on Thursday June 18, 2026 @11:39AM (#48587086)
Parakeet is the name of a speech to text model from Nvidia. Roughly comparable to whisper from openAI.

It's the model doing the work inside the wrapper that an app provides.

[ Parent | on HN ]
Re:DeepSeek Introduces Vision by rob on Thursday June 18, 2026 @11:44AM (#48587153)
Yep, here's the v2 and v3:

https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2

https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3

It's almost instant on my new M5 Max w/ 36GB of memory, but I used both with Handy on my previous 2019 Intel Mac w/ 16GB memory and was completely surprised at just how fast it was for being on-device! Not instant, but only a couple seconds.

[ Parent | on HN ]
Re:DeepSeek Introduces Vision by dghlsakjg on Thursday June 18, 2026 @01:12PM (#48588454)
I’m using it on an M3 max 32gb, and I’m getting 60-70x realtime for recordings and crazy good accuracy. I can get an hour of audio transcribed in a minute. Similar results from Whisper, but half the speed.

Transcription this good used to cost A LOT, now it rounds down to free.

[ Parent | on HN ]