NVIDIA Parakeet

“What surprised me is that GPU is basically sitting idle while Parakeet runs on CPU”

“Parakeet really is quite amazing, I've even had it transcribe words before I've finished saying them!”

“have found pretty good results with some of the smaller models like Parakeet-TDT 0.6B”

“found parakeet V3 tdt int8 onnx very accurate so far and have been pleased with the model”

“I haven't seen anything that has a better speed/accuracy tradeoff when talking to AIs”

“Parakeet did not perform well for me (English with heavy accent).”

“For English I've found Parakeet v2 and Parakeet unified to perform significantly better than v3.”

“Parakeet v2 and Parakeet unified to perform significantly better than v3.”

“they just don't work as well for me, even though they are unquestionably faster.”

“it misses the business specific words when in different language”

“Parakeet has great mandarin>CN text, but running that + a translation model has been tricky and I never got it fast.”

“Nvidia's Parakeet v3 is good enough for my own just local transcription most of the time.”

“It makes a lot more mistakes, possibly because (like you mentioned) large portions of the speech will pause / stutter”

“Recent anecdotal testing I did found Whisper is still better than Parakeet.”

“It always transcribes speech fragments for me, like if I stutter and say "m-m-m-map" parakeet will dutifully transcribe "m m m map"”

“far more accurate in real life dictation compared to even the parakeet models”

“there are better models that are state of the art now like Nemotron and Parakeet”

“Parakeet is way faster (on Nvidia hardware) but not quite as accurate in my experience.”

“You can get amazing local STT using parakeet which can use as little as 600mb of vram”

“Transcription needs to be near-instant, and accuracy "good enough" so that the AI's can "read between the lines". Parakeet-V3 gives exactly this.”

“Parakeet and Whisper are much smaller and perform great for quick dictation and transcription of longer files”

“I found NVIDIA Parakeet to be way better than Whisper - faster, uses less compute, the output is better”

“it will occasionally do this thing where it just repeats a word like a dozen times”

“extremely fast and solid models for their relatively modest memory requirements.”

“Parakeet TDT 0.6B has ~1.9% WER, that's very close to production dictation quality.”

“Google's default speech to text is still bad compared to Whisper and Parakeet”

“Parakeet TDT V2 (English) / V3 (Multilingual) are quite impressive and very fast”

“Thus it saves space (only using about 600MB) while maintaining about the same level of accuracy.”

“for my purposes with Handy I think I need the extra speed and accuracy Parakeet v3 is giving me”

“parakeet v3 has a much better RTFx than moonshine, it's not just about parameter numbers. Runs faster.”

“The voice layer (Parakeet + Kokoro) runs locally on the M4 so that's zero cost.”

“transcription is extremely fast (near instant for a couple sentences, 1-3 seconds for long ramblings)”

“so reverted to v2 for my usage”

“I've noticed large blocks occasionally dropped as well, so reverted to v2 for my usage”

“That has been my go-to model locally and it's hard to imagine there's something even better.”

“Parakeet is pretty limited in the number of languages it supports”

“all of the state-of-the-art speech-to-text systems are large-language model based and are significantly faster and better than what we had before”

“It's better than Whisper, and faster, while running on CPU on my ten year old ThinkPad.”

“I'm running Parakeet V3 on my CPU, on a ten year old laptop, and it works great.”

“What it manages to achieve locally without hogging too much resources is awesome”