Moonshine Open-Weights STT models
AI
speech to text
SHOW HN2026
316points
81comments
#135of 4,366 in 2026
Top 4%of 2026 by points
How HN took it
MixedNo clear lean
20 positive, 22 neutral or asking, 1 mixed, 8 critical, out of 51 comments about the project. Thread score +0.12.
Performance: 15 for, 3 againstIdea: 4 for, 0 againstDocs: 0 for, 3 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Streaming latency is very fast, including on CPU and M1 hardware.
From 4 comments
Streaming transcription is crazy fast on an M1. Would be great to use this as a local option versus Wispr Flow.
saltwounds on HNI've been using Moonshine since V1 and the results are really great. I'd say on par with Parakeet V3 while working really well with CPU only.
dSebastien on HNThis is awesome, well done guys, I’m gonna try it as my ASR component on the local voice assistant I’ve been building https://github.com/acatovic/ova. The tiny streaming…
armcat on HNThe streaming latency numbers are what stand out to me here. Accuracy benchmarks get all the attention, but for real-time applications (voice assistants, live captioning, in-call…
aplomb1026 on HN
Competitive with Parakeet V3 for accuracy on CPU.
From 2 comments
I've been using Moonshine since V1 and the results are really great. I'd say on par with Parakeet V3 while working really well with CPU only.
dSebastien on HNI think there is a streaming version of Parakeet. It is often referred to as Nemotron, though. I tried comparing Parakeet streaming with Moonshine streaming. Moonshine is…
Leftium on HN
Objections
Accuracy is weaker than Parakeet V3 in some tests and on streaming English.
From 3 comments
According to the OpenASR Leaderboard [1], looks like Parakeet V2/V3 and Canary-Qwen (a Qwen finetune) handily beat Moonshine. All 3 models are open, but Parakeet is the smallest…
Karrot_Kream on HNThat was my experience when I tried Moonshine against Parakeet v3 via Handy. Moonshine was noticeably slower on my 2018-era Intel i7 PC, and didn't seem as accurate either. I'm…
SyneRyder on HNImplemented this to transcribe voice chat in a project and the streaming accuracy in English on this was unusable, even with the medium streaming model.
starkparker on HN
Non-English models are under a non-commercial license and only English weights are open.
From 1 comment
> Models for other languages are released under the Moonshine Community License, which is a non-commercial license. Weird to only release English as open weights.
RobotToaster on HN
Claimed accuracy lacks language breakdown and clear benchmark details.
From 2 comments
Accuracy is often presumed to be english, which is fine, but it's a vague thing to say "higher" because does it mean higher in English only? Higher in some subset of languages?…
fareesh on HNClaiming higher accuracy than Whisper Large v3 is a bold opening move. Does your evaluation account for Whisper's notorious hallucination loops during silences (the classic…
heftykoo on HN
Requests
Add model size to the leaderboard comparison.
From 1 comment
Parakeet V3 is over twice the parameter count of Moonshine Medium (600m vs 245m), so it's not an apples to apples comparison. I'm actually a little surprised they haven't added…
reitzensteinm on HN
Support for browser, WebAssembly or ONNX.
From 4 comments
onnx models for browser possible?
sroussey on HNVery cool. Anyway to run this in Web assembly, I have a project in mind
999900000999 on HNThe Parakeet vs Moonshine debate here is missing the point imo. For anything server-side or desktop, sure, Parakeet wins on raw WER. But Moonshine at 245M params vs 600M+ is a…
Paddyz on HNIf it's using ONNX, can this be ported to Transformers.js?
fudged71 on HN
Include a LICENSE file for all code.
From 2 comments
No LICENSE no go
cyanydeez on HNThe LICENSE file that refers to is missing. There's one in the python folder, but not for the rest of the code.
mkl on HN
Questions
How does it compare to Parakeet?
From 2 comments
How does this compare to Parakeet, which runs wonderfully on CPU?
g-mork on HNAccording to the OpenASR Leaderboard [1], looks like Parakeet V2/V3 and Canary-Qwen (a Qwen finetune) handily beat Moonshine. All 3 models are open, but Parakeet is the smallest…
Karrot_Kream on HN
Does it run on Raspberry Pi or phones?
From 2 comments
Oh this is fantastic. I'm most interested to see if this reaches down to the raspberry pi zero 2, because that's a whole new ballgame if it does.
regularfry on HNwhat is the system requirement? can it run on CPU only ? phones ? edit: This is actually amazing. I'm using handy and parakeet to type this message with just my voice. Wow.
agentifysh on HN
What are the streaming latency and partial stability metrics?
From 2 comments
Congrats on the results. The streaming aspect is what I find most exciting here. I built a macOS dictation app (https://github.com/T0mSIlver/localvoxtral) on top of Voxtral…
T0mSIlver on HNNice work. One metric I’d really like to see for streaming use cases is partial stability, not just final WER. For voice agents, the painful failure mode is partials getting…
guerython on HN
Alternatives named
Parakeet V3
From 2 comments
According to the OpenASR Leaderboard [1], looks like Parakeet V2/V3 and Canary-Qwen (a Qwen finetune) handily beat Moonshine. All 3 models are open, but Parakeet is the smallest…
Karrot_Kream on HNParakeet V3 is over twice the parameter count of Moonshine Medium (600m vs 245m), so it's not an apples to apples comparison. I'm actually a little surprised they haven't added…
reitzensteinm on HN
Whisper
From 2 comments
I've helped many Twitch streamers set up https://github.com/royshil/obs-localvocal to plug transcription & translation into their streams, mainly for German audio to English…
francislavoie on HNClaiming higher accuracy than Whisper Large v3 is a bold opening move. Does your evaluation account for Whisper's notorious hallucination loops during silences (the classic…
heftykoo on HN
Handy
From 1 comment
Handy is amazing. Super quality app.
tuananh on HN