Live data from Hacker News

Show HN: Moonshine Open-Weights STT models – higher accuracy than WhisperLargev3

github.com

91–92 of 92 posts

Re: Show HN: Moonshine Open-Weights STT models – higher accuracy than WhisperLargev3

#91

Earlier quoted context omitted.

parakeet v3 has a much better RTFx than moonshine, it's not just about parameter numbers. Runs faster. https://huggingface.co/spaces/hf-audio/open_asr_leaderboard

It is about the parameter numbers if what you care about is edge devices with limited RAM. Beyond a certain size your model just doesn't fit, it doesn't matter how good it is - you still can't run it.

I am not sure what "edge" device you want to run this on, but you can compress parakeet to under 500MB on RAM / disk with dynamic quants on-the-fly dequantization (GGUF or CoreML centroid palettization style). And retain essentially almost all accuracy.

And just to be clear, 500MB is even enough for a raspberry Pi. Then your problem is not memory, is FLOPS. It might run real-time in a RPi 5, since it has around 50 GFLOPS of FP32, i.e. 100 GFLOPS of FP16. So about 20-50 times less than a modern iPhone. I don't think it will be able to keep it real time, TBF, but close.

regardless, this model with such quantization strategy runs real time at +10x real-time factor even in 6-year old iPhones (which you can acquire for under $200) and offline at a reasonable speed, essentially anywhere.

You get the best of both worlds: the accuracy of a whisper transformer at the speed and footprint of a small model.

Re: Show HN: Moonshine Open-Weights STT models – higher accuracy than WhisperLargev3

#92

Earlier quoted context omitted.

Parakeet doesn't require a GPU. I'm handily running it on my Ubuntu Linux laptop.

you are right i just downloaded it on handy and its working i can't believe it i was using assmeblyAI but this is fast and accurate and offline wtf!

parakeet is amazing, it has completely ousted whisper for me. On Linux, both handy.computer and epicenter Whispering (using parakeet of course) work incredibly well for set-and-forget STT. I use it constantly to write messages on Slack/Teams, do debate with claude code etc. Both have minor bugs, but I can easily accept those, these apps being FOSS and all.

On Mac, I've been using VoiceInk and it's even better. VoiceInk (and MacWhisper too, IIRC) use the neural engine and the delay between dictation and appearance of the typed text is almost imperceptible.

Post reply on HN