Whisper is the wrong model to benchmark against, or rather, there are better models that are state of the art now like Nemotron and Parakeet both by Nvidia, as well as Mistral's Voxtral and Cohere Transcribe.
However, what's funny is, RIP to a lot of the paid apps that simply wrap Whisper, I'm sure Apple will make a native GUI such as a recorder app for macOS that obviates the need for these wrappers, which everyone seems to be vibe coding these days.
Every single asr model I tested so far did not support timestamps properly though. Some use external aligner to create timestamp, but the accuracy is still much inferior than whipser in case the audio is noisy.
Just ran it against Whisper-Large-V2 on a math lecture (my primary use case for ASR is subtitling math lectures), and it was substantially faster and only slightly worse. Very usable for live transcription though I'll probably stick with whisper for the time being since I don't really need the subtitles to be generated in real time.
Been using it for a podcast app I have been developing for half a year lol (I hope I publish it by version 27) and I can confirm it’s real fast.
Splitting the audio in multiple segments and firing it up without hitting the maximum limit of concurrent decoding streams makes it blazing fast. Fair enough you loose the cut, but it’s good enough for just podcast. In one minute it chews through one hour of audio. This on an iPhone 17 Pro.
What's different about your podcast app?
Nothing really, except that I get to play with SpeechAnalyzer APIs, foundation models, translations. It’s basically my playground where to try all things. Been listening a lot of Chinese podcasts lately, transcribed and translated by local models.
Edit: all that said, the app is irrelevant. What I want to say is that live transcripts on iOS using Apples frameworks works very well. Only thing I miss is diarization support.
If it was faster but worse, maybe compare it to a smaller whisper model?
I imagine because quality of transcription is what matters.
Whisper small/tiny/base are almost four years old (they were not updated for Whisper v2 or v3). Is there really nothing better to benchmark against by now?
There are many [0], you can search and filter by streaming and open weight only as well.
Looks like Voxtral and Nvidia's Nemotron are best.
[0] https://artificialanalysis.ai/speech-to-text/non-streaming
There's tons, Parakeet was the last I remember seeing which seemed to gain traction (independent lightweight implementations etc).
I have tried everything (that will run on a 12GB RTX 4070) and I have yet to find anything with better accuracy than Whisper V2 Large for my dataset (discord audio from TTRPG sessions, isolated per-speaker, mostly non-American accents)
Same, for my English-only podcast
Nvidia's Nemotron subsumes their older Parakeet model now even for real time streaming.
Parakeet is way faster (on Nvidia hardware) but not quite as accurate in my experience.
It's also super fast on CPU.
Parakeet isn't as good as whisper large.
For my current purposes, I need a speech-to-text model/API to also emit word-level timestamps - for now, that makes ElevenLabs's Scribe v2 the best multiplatform, multi-language choice though it does look like this SpeechAnalyzer API provides them (although only for English).
Impressive. Apple said they improved the models in 27 didn’t they? It would be interesting to see the numbers the beta turns in.
Any chance you can benchmark against whisper large and large v3 turbo? These run comfortably on older Macbooks and are still far more accurate in real life dictation compared to even the parakeet models( despite ASR leaderboards) with an RTF < 1.
Try MOSS-Transcribe-Diarize from a few days ago. I’m getting better results than those whisper models. And it’s very fast and small. Better suited to noisy audio too.
I use Spokenly, offline-only mode with the Nvidia model. All local, totally free. Highly recommend
I second that! Can you run your benchmarks against the iOS 27 beta?
Lots of comments are "you should compare against X and Y" - even better, just get the results on a standard benchmark, so you can compare against all,
well, it's nice, but the multi-lingual diff is rather limited (only European languages).
Vs Voxtral would be a better comparison. No other model, open or closed, has been able to hit such a low AER (Acronym Error Rate ;)) for my meeting transcripts. Seems to understand/infer all the technobabble I use at work. Never have to edit anything. Whisper was catastrophically bad.
I typically disable autocorrect on Apple products because of this, cautiously optimistic about their improved speech models, but definitely worried that it's going to 'correct' technical jargon to more common words.
I stopped reading after seeing they compared only with Whisper Small, Base, Tiny
This is useless test and benchmark when you have these day Whisper-V3-Large and Whisper V3-Turbo that you can faster than realtime on 5 years old macbook on apple sillicon (ANE). They didn't even compared to parakeet v2 or parakeet v3. And only english language...
This is great marketing, I had no idea what inscribe was, but a blog like this going viral did something no ad could do for me.
If this isn't open source/weights and can't run locally, I don't see how this is a replacement for Whisper or other open models, e.g. within Home Assistant.
It's a local model so it's essentially open weight such that you could feasibly export it somehow since it's already on the laptop somewhere. Apfel is a wrapper app like ChatGPT but using Apple Foundation Models, I assume something similar will happen with this transcription model.
It's not open weight, but the point is to be an on device (and thus local, privacy preserving) option. The article mentions that as the caveat
> What this means if you just want good transcription
> If you are on a current iPhone or Mac, the best on-device transcription engine for English is already in the operating system, and the private option is no longer the compromise option
> It's not open weight, but the point is to be an on device (and thus local, privacy preserving) option.
How can you be sure this isn't leaking data or metadata to Apple? Can Apple really be trusted?
Test it! Does it make network requests? Unplug the internet and see if it still works!
It is local
The appeal is that users only have to download it once across all apps that use it. Instead of convincing a user to give a couple gigs for just your one app
Anyone know the best choice these days specifically for speaker diarization?
I'd be interested too. Last time I looked the state of the art was pretty bad.
Yeah i do find apple's speech to text very good lately and no need to use openai or anything that seem to market their services better
Im hoping Apple gets the new Siri working better on older phones. I was excited to use it but the latest beta / Siri runs too slow on my iPhone Pro Max 15.
Im looking for the same experience I have when talking to chatGPT. As for past two years or more talking to GPT within it's app and on my iPhone Pro Max 15 it runs smooth as butter :-). This is the experience I was and still am hoping with Apple, but Im thinking all the extra layers of privacy and security might be slowing them down?
Overall, Apple who is suing Open AI should just buy them and let me have the best conversational AI out there baked into my old ass iPhone. Because as so far the new Siri on my old phone (tho again GPT works great talking to it and for years) doesnt come close. It's the same old "Could you try that again," Siri. BOO!!!
Yeah, ChatGPT voice is great experience vs. Siri on that phone. In case you haven't done something like this already:
1. In Shortcuts app, make shortcut named "AI Voice Mode" (or whatever you want, YMMV)
2. Set it to run the ChatGPT action "Voice Mode" (requires at least the minimum paid tier, I think)
3. To trigger, say "Hey Siri, AI Voice Mode" (or whatever you called the shortcut)
You can also map the Action Button to GPT voice mode.
Ridiculous that for past many years we can talk to GPT on our iPhones without any hiccups and this new Siri is still the same old horse crap (at least for me and this latest beta). Buy them already Apple or possibly be replaced by them as their path & trajectory (working on Ai devices now and they are stealing like Jobs did with Xerox) mirrors yours in the late 1970s.
Thanks for the tip and if Im not mistaken it's similar to asking Siri to ask chatGPT to ask XYZ?
> similar to asking Siri to ask chatGPT to ask XYZ
Effectively, it sort of does that, but really it just listens to the wakeword and opens/switches to the requested app & modality.
FWIW, I get a very different functional result using the Shortcut method vs. asking Siri to delegate natively. To compare, I asked Siri (non-beta here) now to "ask ChatGPT <x>" and I got a top-card with some fairly low quality SEO-ranked weblinks.
I’m on iPhone 17 Pro Max, 27 beta 3.
New Siri is impressive in that it answers satisfactorily now 80% of the time vs 10% with old Siri.
But it’s slow as shit. GPT, Claude, and Gemini can answer me in 5-10 seconds. Google AI Mode can answer in 2 seconds.
New Siri usually takes 25 seconds to respond to me. This morning it timed out (with strong network connection) when asked a simple multiplication question.
Damn slow on your newer phone too. Pathetic Apple!
> Im hoping Apple gets the new Siri working better on older phones.
Apple would never do that, if anything they did not offer their Siri with the most advanced AI on iPhone 16 Pro Max, which is one year-old only.
Author here. I ship both Apple speech engines plus WhisperKit side by side in a transcription app, which made it possible to run all five through identical production code on the same audio: LibriSpeech test-clean and test-other, 5,559 utterances, fully on-device on an M2 Pro.
Apple published no accuracy numbers for SpeechAnalyzer (or for SFSpeechRecognizer, ever, as far as I can tell), so the migration question has been guesswork. Short version: the new API cuts WER 3.5-4x vs the old one (2.12% vs 9.02% on test-clean), and it also beat Whisper Small on both splits at about 3x the speed. The old API came in last on clean speech, behind even Whisper Tiny.
On "why should I trust a vendor benchmark": the Whisper column reproduces OpenAI's published LibriSpeech WERs within +0.11 to +0.42 on all six measurements (same corpus, same normalizer, same scorer for every engine), and the raw per-utterance transcripts are downloadable from the article if anyone wants to rescore with their own normalizer.
Limitations worth stating up front: English only, read speech rather than meeting audio, one machine. Precise per-engine timing isn't in the article yet because the accuracy runs shared the machine with a dev workload; WER is load-independent, timing isn't.
Two things that might interest people migrating: SFSpeechRecognizer sends audio to Apple's servers unless you set requiresOnDeviceRecognition, and with SpeechAnalyzer, finishing your input stream is not enough to end a session. If you never call finalizeAndFinishThroughEndOfInput(), the results sequence never terminates and your await hangs forever. I found that one because it was shipping in my own app.
Happy to answer questions about the harness or the normalizer.
At this point, I would not recommend ignoring Parakeet TDT 0.6b v2/v3 (english-only versus multilingual). Those models have been out for a year, give or take, and they are both accurate and fast. I would choose Parakeet over Whisper in almost all situations these days. Parakeet works great even on my several year old iPhone 15 Pro Max, so if an app is going to ship a dedicated model, I strongly recommend investigating Parakeet.
On the more cutting edge front, Granite Speech 4.1 has proven to be a reliable workhorse for me, but it is larger than Parakeet. Cohere Transcribe is interesting, but how strong it is seems to vary more from task to task.
Parakeet Unified 0.6B came out a few months ago, combining both online streaming and offline transcription into one model, and that is one that I need to test more, but it seems promising.
As others have mentioned macOS 27/iOS 27 is supposed to have a new model, particularly on devices with 12GB of RAM or more. I have not actually seen the option to enable that new model yet, though, despite being on the beta on a device that meets the requirements. Maybe a benchmark would reveal that it is already active?
Why do you not use Whisper large models when on macOS? They're still fast even when streaming and yield a much lower WER.
Also, just out of curiosity, seems like everyone and their mother is making Whisper wrappers, how is your app different?
Why use relatively ancient models like whisper and not e. g. parakeet?
Please run your benchmark on this new and very impressive model https://huggingface.co/OpenMOSS-Team/MOSS-Transcribe-Diarize in my testing it outperforms all mentioned especially on noisy audio
MOSS-Transcribe-Diarize
Still nothing beats OpenAI's VTT. Anthropic's sucks and Apple's isn't even usable.
Edit: Getting downvoted by Apple fanboys for telling the truth is a badge of honor.
Which OpenAI model/API do you mean?
Whisper and GPT-4o (for diarization).
In which world 98% accuracy is not usable?
In a world where you say "tmux" and Apple's VTT writes "T Max".
You can't reasonably expect generic ASR to infer tmux from "tee-mucks". "tee-em-you-ex" works reliably if you're ok with capitalisation for your use case.
agreed. plus all the languages supported!
this is amazing. if i had a mac i would try to reverse engineer the code, extract the weights and port it to something that works on linux/windows like torch or burn. then put the code on github and weights on a torrent site. lifes too short to let apple keep their models exclusive.
Probably not worth the effort (or legal trouble), unless you can show it's better than other recent open models like Cohere Transcribe.
Is that copyright infringement?
This hasn't been tested in court. But there's a high chance that model weights are not copyrightable, only the code to generate them is.
Cloud models are usually protected by trade secret laws, leaking them would get you in trouble. However if the model is made available publicly, as long as you don't break the law to get them, anything after that would be fair game unless Apple can prove that humans have significant authorship over the weights, which hasn't been tested and is a significant burden to prove/disprove.
Copyright protects original forms of expression, not arbitrary data. It is very arguable whether it applies to model weights. However, it would likely constitute a license violation.
Aside from the legality of it, I think you are underestimating how complex it can be to do that. It is possible in theory but not something that will be a fun side quest like you are making it seem.
With a IDA Pro decompiler license & MCP server, paired with Codex/Claude Code... it would be a fun side quest.
You likely don't need to disassemble the inference code, the weights are "just an array of numbers" in MLX format.
This.
The Jedi Hand Wave-y nature of the way people talk about AI these days is going to make reigning in the AI superpowers nearly impossible. Because there are people out here who believe models of this quality are easily replicated or reverse engineered. Neither is really doable on any reasonable timeline by people who are not AI experts. Real AI experts. Not TF/PyTorch monkeys or Agent Slop Slingers.
And those people are already highly incentivized to not make anything performing better than SOTA models open source.
[dead]
Also, this test is English-only, while a strong point of other models is to understand different languages without first having to say which one (so you don't need 3 different keyboard shortcuts if you wanna dictate in 3 languages day-to-day)
As an Australian, Apples voice models have always sucked. I've tried using stt (again) more recently and its improved, but i'm so tired of having to Americanize my voice to get it to figure out what the hell i'm saying.
As a Texan first, American second, I sympathize with this statement. Siri can't understand me probably 25% of the time. I use STT for iMessage while in the car, and half the time it will take 3+ times to either get it right or me give up, and hope to remember to text them by hand when I next stop.
It also struggles with my NYC-area accent, which is only medium thick.
> there are better models that are state of the art now like Nemotron and Parakeet both by Nvidia
Is parakeet state of the art? 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". Which I guess could be a good thing or a bad thing depending on what you want. Whisper does not do that however.
I do like cohere transcribe a lot.
Agree on this point. Recent anecdotal testing I did found Whisper is still better than Parakeet.
Apple's own Voice Memos app already does automatic transcription since macOS 15 / iOS 18.
Speech-to-text is also already built into the keyboard as well, so it can be used in any app where a user would type.
From my experience, Speech-to-text falls way short of Wispr flow and I would assume the ones that are said to be better than that. It lacks context awareness and formatting
For multilingual and noisy audio the best right now is MOSS-Transcribe-Diarize which was released just a few days ago
Superwhisper does a lot more than just provide a whisper/parakeet UI so I’m not sure Apple will destroy them so easily
Interesting! And what would you say are MTD top competitors?
Thanks, was looking at a better diarization model.
Even for those sorts of apps, MacParakeet which I've been using is FOSS so no payment needed. In reality these days with AI the ability to spin up a free and/or OSS competitor falls to zero.
I’m not even using it for diarisation just transcription and it’s amazing. It also doesn’t need a VAD
A new VAD I found though is FireRedVAD which has better benchmark results than TEN and Silero by far
Came here to post this. I use handy on my own machine and it's perfect with parakeet. If I switch to whisper it makes lots of mistakes