I’ve tested at least 20 STT models in a benchmark I’ve set up with German, Italian and English voices from meetings in my company.
The voices contained very industry specific words, the languages changed from one sentence to another, sometimes words in a language were mentioned while a discussion was in another.
The only local model that satisfies me is Voxtral Mini 3b, the only paid API that is slightly better is eleven labs. Yes, Voxtral might not reach the best score in the benchmarks, but to me, it just solves a problem. It might not be the best in terms of speed... but that’s not a problem for me.
Happy to test this new model from Google but I’m not sure I’d go with that instead of something that can run so easily in my machine.
It's interesting that the results can be so different depending on the person, the use case and even the microphone used. I use dictation a lot, so I try to stay up to date with the latest models as much as I can. So far, for my needs, nothing could beat Whisper Large v3. I keep hearing that Nvidia Parakeet models are better, but they just don't work as well for me, even though they are unquestionably faster.
Things change a lot if you need to speak to the model in multiple languages. There are very few models out there that can automatically detect the language spoken and produce correct output.
I tried the larger Voxtral models, but they didn't work for me at all. When I spoke to them in Polish, they produced output in Russian or Ukrainian.
For now, I settled on creating my own plugin for TypeWhisper, which runs Whisper Large on the GPU and does it much faster than pretty much anything else out there. But I'm still hoping that something better will come along, as Whisper Large is quite old at this point.
Agree about Voxtral being the only model — local or cloud — that handles multilingual conversation really well. I‘m not sure what they do differently, but mixed-language sentences and industry terms don‘t seem to faze it where other model begin to struggle.
What i want is a model that outputs its predictions and their scores along with the text it choose. So I could flag something it's getting consistently wrong, like mis-predicting a technical term or name, or acronym, and say replace it with my correction, and have the ui be able to smartly replace that in the whole text so far and future parts. even better would be the ability to feed this back into the model for future runs.
Good news: most of these models can include a prompt that steers the transcription; if you use frequently a word you just invented, add it there and it will be transcribed correctly more likely.
Do you use it on a desktop? Mac by any chance? What's your setup?
I've been looking to find a simple and fast dictation app for English but almost everything I've tried (from Handy to many apps, eg some with Whisper in their names, after the model I assume) just don't work well. Apple's offering is worse than those though. I even tried with local enhancement models.
I am biased as a developer on the project but you should give Epilude [1] a go if you want local dictation on a Mac.
We’ve fine-tuned open-weight models to make them better (in our benchmarks) at cleaning up and formatting what you say so you don’t have to edit what you dictate.
This problem is legitimately hard and needs high cognitive abilities. Even the biggest generalist models struggle with memes and lingo salad that sound immediately intuitive for an out-of-the-loop human, and I'm talking about text comprehension. Modern models are optimized for decision making and are worse in that than old ones optimized for creative writing, but those also struggled.
You should probably not expect a small STT model like Voxtral or Parakeet to do any better, unless it's laser focused on that area in particular and sucks in everything else.
I love it but it misses the business specific words when in different language. Sometimes it stretches them out to fit unrelated words in the language of the rest of the conversation. I miss its speed though.
Of course not, but it’s easily replicable just by mixing different languages conversations together, adding a word here and there of some very specific German jargon.
"Function calling: The model can delegate complex tasks (such as image generation and file analysis) to other Gemini models via function calls. Currently available in the Gemini macOS app."
This confused the heck out of me because it makes it sound like the STT model can make function calls in order to execute arbitrary tasks, which wouldn't make any sense. The developer docs (https://ai.google.dev/gemini-api/docs/models/gemini-3.5-tran...) confirm that the Gemini 3.5 Transcribe model cannot in fact make function calls. I guess the blog post is just using very confusing wording to describe how their consumer assistant/chatbot app can both take audio input via Gemini 3.5 Transcribe and then call other things as needed. Or maybe that bullet point was meant to be for a different model announcement and somebody made an editing error.
Maybe I missed something, but isn’t it impossible to detect whether something is written by AI or not? A human on a bad day can write like AI while an AI on a good day can write like a human.
No, that's not correct for any reasonable definition of "impossible." Look up pangram's accuracy ratings. It's not perfect, but it's pretty good. LLMs in fact leave very distinguishing traces of their logit distributions in the text they write. It's one of the reasons why it's so easy for humans to also smell them.
It is possible to trick pangram - they bias toward a low false positive and a higher false negative - but it is not true that it is essentially random.
Remember that function calling doesn’t mean that the model is actually calling a function for you: what it does it interpret the intent, then return back to the application the function and arguments, then the application runs the function with those arguments.
Think voice control for your device. You speak, and it returns back instructions ( a function and arguments) for your device to be execute.
STT models just turn audio into text, though; they don't interpret what the text means or figure out how to translate it into other representations like function calls. You would use a general-purpose model for that.
I've been testing it on Pixel 11 Pro and I mostly dislike it. It is convenient when you have something long to say without thinking about it first.
But the main issue is when you want to say something precise with specific wording it might "simplify" it and break the meaning.
Something like "I hesitated to check it, I should have verified" => "I should have verified" (The "I hesitated..." is removed but I said it because I wanted to let the person know that I thought about it earlier)
I tried the sentence few times and it always removed the first part.
I personally tested all the STT models for my real-time translator (https://fliptalk.ai). From language detection and accuracy in a noisy environment to the most important point: latency.
At the moment, Soniox STT v5 is definitely the best, and I'm impressed by its performance. It's good that Google released Gemini-3.5-Transcribe, and it beats every other model on accuracy, but it definitely needs a bit more work on latency, which is the most important factor for STT apps.
For me, Gemini-3.5-Transcribe actually has slightly lower latency. Kudos to Soniox for both paying their competitor and letting them win. But yes, Soniox is much cheaper.
Oh, thanks for pointing me to Soniox. It is really good. Would also pick up the words with different languages, identify and output in the right language.
Looks interesting. It was much faster too, but that I cannot say much since it was on their own website.
Interesting, I uploaded a voice recording from a meeting I had recorded with a relatively cheap microphone.
Soniox came out really good. OpenAI started getting some things very wrong and even introduced some German. Google did okay but cut off the start by several seconds.
What's Soniox doing (left most) that's making it so good ? It was also the only one that could distinguish between the speakers.
> latency, which is the most important factor for STT apps.
Perhaps latency is more important than accuracy for a real time translation app (I actually disagree with this - imagine e.g. the hilarity when requesting "a new display" being translated as "a nudist play"), but certainly not for all applications. My pet app transcribes personal voice notes to self, it could run all night.
Realtime + Voice AI usecases is where latency is most important. I use Handy on my desktop and i can tolerate a latency of a few seconds every now and then. Your P99 should on TTFB should be really low to compete for voice ai realtime
I'd certainly use Gemini API a lot more if I could figure out how to give them money and they give me tokens. The whole tier nonsense is just insane hoop jumping compared to their competition.
I've been using Wispr Flow on my pixel 9 and I like it. It does the things that Rambler says it's supposed to do but it's available on my Pixel 9 unlike rambler that's currently only available for pixel 11.
Curious if this suffers from the same hallucination problems as Chirp or not. For example, if you pass chirp some audio with noise or even no audio, it will barf text at you like "I don't know. I don't know. I don't know." until a request timeout fires after like 10 minutes. It's... really bad. For this reason, we've moved back to Whisper for timestamp accuracy and then Gemini Flash for transcription correction.
I eval'd this pretty heavily with no audio or just noise, as for us hallucinating a response is very bad. It works great in this scenario. There are some bugs, e.g. randomly exactly 20 seconds of silence will lead to a 403, but I'm sure these aspects will be improved over time. I didn't get any hallucinations though.
Yes, that's sort of our strategy as well: for us accuracy (and following a style guide, and extracting on-screen text from video) is important. So we do a first pass with whisper to get the word timestamps, and the pass the same audio (chunked in case of longer ones) to Gemini (pro, not this model), to transcribe without timestamps. After that you can transfer the timestamps from whisper to the timestamp-less Gemini text, by looking for runs of identical words in both and fudging timestamps for the other words I'm between a bit. Works quite well, and I'll be tesing this model to see if it can replace Whisper.
I've seen Whisper do this exact thing as well. And also repeating a few words over and over near the end, words that were said but not completely at the end. It seems to overwrite the last few sentences with that so actual content is lost. It's really annoying especially because Whisper is really quite good otherwise. Sometimes a friend sends me audio messages from a noisy car and I can't make it all out but whisper gets it mostly right.
I’ve noticed the some of YouTube’s auto-generated captions will sometimes hallucinate “Heat” during long music segments; probably an older model though.
That's really impressive! It says it's available in GBoard on Android but I don't see it yet. I assume like most Google features it will be a rolling release over the course of several months, can't wait to try it some day.
It is confusing. As far as I can tell "rambler for Android" is a named subfeature of gboard and only on Pixel 10+ and newer Samsung phones. From a user standpoint it "just" seems like the transcription has gotten better - not an app you use or a feature that you enable.
I would also like to see benchmark for translation. I'm looking for live translated subtitles so my Japanese wife can enjoy any show with out waiting months for official VOD streams to release them.
Still no real time diarization beyond 3 people (and even then experimental) when others do it very well, like Soniox and Deepgram. For something like meeting notes this is critical. Not sure what the issue is to implement it, maybe that's not Google's use case in mind and rather it's about personal Rambling as the feature on Pixels shows, which uses this model.
Yes, I specifically avoided mentioning that because they don't need to diarize from audio directly as they know exactly which participant is speaking at all times.
I am not sure if "Word Error Rate" captures what has always been wrong with transcription. My biggest complaint is that it inserts sentence breaks in random places, then fails to evaluate the result, even though it is obviously wrong. Then I have to go fix it which can be harder than having just typed it myself, due to the difficulty of positioning the Android cursor, the fact that it automatically capitalizes if you delete a capital letter, etc. And much of the time I fail to notice the errors until later.
have another model do a pass to clean it up, saw a demo of local STT where someone did this, can fix a lot of things, especially with gotchas for the STT model in a clean-transcript.md
I think the model can even evaluate itself. If it looks afterward at an output like "do you. Want to get lunch?" in the absence of affirmative evidence that the user wanted it that way, it should be able to see that it goofed.
I'd love to know how this handles proper subtitle formatting. I'm in the process of learning many languages, and being able to cross check my own understanding with film and video would be fantastic.
Google is crushing it! They’re going toe to toe against the European giant Mistral with all these small useful models and they have a chance of winning! Never have I been more excited about Google’s capabilities.
I’ve tested at least 20 STT models in a benchmark I’ve set up with German, Italian and English voices from meetings in my company.
The voices contained very industry specific words, the languages changed from one sentence to another, sometimes words in a language were mentioned while a discussion was in another.
The only local model that satisfies me is Voxtral Mini 3b, the only paid API that is slightly better is eleven labs. Yes, Voxtral might not reach the best score in the benchmarks, but to me, it just solves a problem. It might not be the best in terms of speed... but that’s not a problem for me.
Happy to test this new model from Google but I’m not sure I’d go with that instead of something that can run so easily in my machine.
It's interesting that the results can be so different depending on the person, the use case and even the microphone used. I use dictation a lot, so I try to stay up to date with the latest models as much as I can. So far, for my needs, nothing could beat Whisper Large v3. I keep hearing that Nvidia Parakeet models are better, but they just don't work as well for me, even though they are unquestionably faster.
Things change a lot if you need to speak to the model in multiple languages. There are very few models out there that can automatically detect the language spoken and produce correct output.
I tried the larger Voxtral models, but they didn't work for me at all. When I spoke to them in Polish, they produced output in Russian or Ukrainian.
For now, I settled on creating my own plugin for TypeWhisper, which runs Whisper Large on the GPU and does it much faster than pretty much anything else out there. But I'm still hoping that something better will come along, as Whisper Large is quite old at this point.
Agree about Voxtral being the only model — local or cloud — that handles multilingual conversation really well. I‘m not sure what they do differently, but mixed-language sentences and industry terms don‘t seem to faze it where other model begin to struggle.
What i want is a model that outputs its predictions and their scores along with the text it choose. So I could flag something it's getting consistently wrong, like mis-predicting a technical term or name, or acronym, and say replace it with my correction, and have the ui be able to smartly replace that in the whole text so far and future parts. even better would be the ability to feed this back into the model for future runs.
Good news: most of these models can include a prompt that steers the transcription; if you use frequently a word you just invented, add it there and it will be transcribed correctly more likely.
Do you use it on a desktop? Mac by any chance? What's your setup?
I've been looking to find a simple and fast dictation app for English but almost everything I've tried (from Handy to many apps, eg some with Whisper in their names, after the model I assume) just don't work well. Apple's offering is worse than those though. I even tried with local enhancement models.
I am biased as a developer on the project but you should give Epilude [1] a go if you want local dictation on a Mac.
We’ve fine-tuned open-weight models to make them better (in our benchmarks) at cleaning up and formatting what you say so you don’t have to edit what you dictate.
https://epilude.com
I used MacWhisper for a long time and then recently switched to the free TypeWhisper. Both work very well on a Mac.
Out of curiosity, what was wrong with Handy? I use it and it works fine.
You should give VoiceInk[1] a go. It works great and has decent latency.
https://github.com/Beingpax/VoiceInk
This problem is legitimately hard and needs high cognitive abilities. Even the biggest generalist models struggle with memes and lingo salad that sound immediately intuitive for an out-of-the-loop human, and I'm talking about text comprehension. Modern models are optimized for decision making and are worse in that than old ones optimized for creative writing, but those also struggled.
You should probably not expect a small STT model like Voxtral or Parakeet to do any better, unless it's laser focused on that area in particular and sucks in everything else.
have you tried groq whisper-turbo? works well for me for multiple different languages
I use Parakeet 3. How does that fare in your testing?
I love it but it misses the business specific words when in different language. Sometimes it stretches them out to fit unrelated words in the language of the rest of the conversation. I miss its speed though.
Any possibility to share some of the eval audio?
Of course not, but it’s easily replicable just by mixing different languages conversations together, adding a word here and there of some very specific German jargon.
Try the recent MOSS one? It’s very good
"Function calling: The model can delegate complex tasks (such as image generation and file analysis) to other Gemini models via function calls. Currently available in the Gemini macOS app."
This confused the heck out of me because it makes it sound like the STT model can make function calls in order to execute arbitrary tasks, which wouldn't make any sense. The developer docs (https://ai.google.dev/gemini-api/docs/models/gemini-3.5-tran...) confirm that the Gemini 3.5 Transcribe model cannot in fact make function calls. I guess the blog post is just using very confusing wording to describe how their consumer assistant/chatbot app can both take audio input via Gemini 3.5 Transcribe and then call other things as needed. Or maybe that bullet point was meant to be for a different model announcement and somebody made an editing error.
Probably written using Gemini which hallucinated.
Pangram says human: https://www.pangram.com/history/a69f9b74-eb46-44b9-a087-7822...
Maybe I missed something, but isn’t it impossible to detect whether something is written by AI or not? A human on a bad day can write like AI while an AI on a good day can write like a human.
No, that's not correct for any reasonable definition of "impossible." Look up pangram's accuracy ratings. It's not perfect, but it's pretty good. LLMs in fact leave very distinguishing traces of their logit distributions in the text they write. It's one of the reasons why it's so easy for humans to also smell them.
It is possible to trick pangram - they bias toward a low false positive and a higher false negative - but it is not true that it is essentially random.
https://bfi.uchicago.edu/wp-content/uploads/2025/09/BFI_WP_2...
Remember that function calling doesn’t mean that the model is actually calling a function for you: what it does it interpret the intent, then return back to the application the function and arguments, then the application runs the function with those arguments.
Think voice control for your device. You speak, and it returns back instructions ( a function and arguments) for your device to be execute.
STT models just turn audio into text, though; they don't interpret what the text means or figure out how to translate it into other representations like function calls. You would use a general-purpose model for that.
I've been testing it on Pixel 11 Pro and I mostly dislike it. It is convenient when you have something long to say without thinking about it first. But the main issue is when you want to say something precise with specific wording it might "simplify" it and break the meaning.
Something like "I hesitated to check it, I should have verified" => "I should have verified" (The "I hesitated..." is removed but I said it because I wanted to let the person know that I thought about it earlier)
I tried the sentence few times and it always removed the first part.
Are you in Smart or Verbatim mode?
https://ai.google.dev/gemini-api/docs/transcribe#transcripti...
They mentioned the pixel phone so probably using https://blog.google/products-and-platforms/platforms/android...
I personally tested all the STT models for my real-time translator (https://fliptalk.ai). From language detection and accuracy in a noisy environment to the most important point: latency.
At the moment, Soniox STT v5 is definitely the best, and I'm impressed by its performance. It's good that Google released Gemini-3.5-Transcribe, and it beats every other model on accuracy, but it definitely needs a bit more work on latency, which is the most important factor for STT apps.
Soniox website has a live comparison demo: https://soniox.com/compare-stt
For me, Gemini-3.5-Transcribe actually has slightly lower latency. Kudos to Soniox for both paying their competitor and letting them win. But yes, Soniox is much cheaper.
Oh, thanks for pointing me to Soniox. It is really good. Would also pick up the words with different languages, identify and output in the right language. Looks interesting. It was much faster too, but that I cannot say much since it was on their own website.
Interesting, I uploaded a voice recording from a meeting I had recorded with a relatively cheap microphone.
Soniox came out really good. OpenAI started getting some things very wrong and even introduced some German. Google did okay but cut off the start by several seconds.
What's Soniox doing (left most) that's making it so good ? It was also the only one that could distinguish between the speakers.
Depends on your use case. If you're transcribing meeting notes, latency is a non-issue.
Realtime + Voice AI usecases is where latency is most important. I use Handy on my desktop and i can tolerate a latency of a few seconds every now and then. Your P99 should on TTFB should be really low to compete for voice ai realtime
I'd certainly use Gemini API a lot more if I could figure out how to give them money and they give me tokens. The whole tier nonsense is just insane hoop jumping compared to their competition.
I use OpenRouter. Interacting with the Google Cloud Console is pure pain.
Its easy to pre-pay tho?
I've been using Wispr Flow on my pixel 9 and I like it. It does the things that Rambler says it's supposed to do but it's available on my Pixel 9 unlike rambler that's currently only available for pixel 11.
I don't really need this model, but I needed a fast my voice to command thing, and Gemini flash 2.5 was extremely effective(low latency, high speed).
No need to go from audio to text to reasoning, just from audio to output json for running a command via adb automatically and it's working crazy good.
Curious if this suffers from the same hallucination problems as Chirp or not. For example, if you pass chirp some audio with noise or even no audio, it will barf text at you like "I don't know. I don't know. I don't know." until a request timeout fires after like 10 minutes. It's... really bad. For this reason, we've moved back to Whisper for timestamp accuracy and then Gemini Flash for transcription correction.
I eval'd this pretty heavily with no audio or just noise, as for us hallucinating a response is very bad. It works great in this scenario. There are some bugs, e.g. randomly exactly 20 seconds of silence will lead to a 403, but I'm sure these aspects will be improved over time. I didn't get any hallucinations though.
Yes, that's sort of our strategy as well: for us accuracy (and following a style guide, and extracting on-screen text from video) is important. So we do a first pass with whisper to get the word timestamps, and the pass the same audio (chunked in case of longer ones) to Gemini (pro, not this model), to transcribe without timestamps. After that you can transfer the timestamps from whisper to the timestamp-less Gemini text, by looking for runs of identical words in both and fudging timestamps for the other words I'm between a bit. Works quite well, and I'll be tesing this model to see if it can replace Whisper.
I've seen Whisper do this exact thing as well. And also repeating a few words over and over near the end, words that were said but not completely at the end. It seems to overwrite the last few sentences with that so actual content is lost. It's really annoying especially because Whisper is really quite good otherwise. Sometimes a friend sends me audio messages from a noisy car and I can't make it all out but whisper gets it mostly right.
I am using it all day as a main dictation model, and I can say it's the best model in terms of accuracy, latency, and formatting I have ever used.
mind if I ask what dictation tool you are using?
I’ve noticed the some of YouTube’s auto-generated captions will sometimes hallucinate “Heat” during long music segments; probably an older model though.
That's really impressive! It says it's available in GBoard on Android but I don't see it yet. I assume like most Google features it will be a rolling release over the course of several months, can't wait to try it some day.
It is confusing. As far as I can tell "rambler for Android" is a named subfeature of gboard and only on Pixel 10+ and newer Samsung phones. From a user standpoint it "just" seems like the transcription has gotten better - not an app you use or a feature that you enable.
Still more expensive and worse performing than ElevenLabs Scribe, unfortunately. Not sure who's the target audience for this.
Just added this to my benchmark site: https://multilingualsttbench.com/
It doesn't reach the frontier in either latency or accuracy for ai multilingual conversations.
missing aqua voice's avalon 1.5 model there.
Thanks for this, really helpful.
I would also like to see benchmark for translation. I'm looking for live translated subtitles so my Japanese wife can enjoy any show with out waiting months for official VOD streams to release them.
I'm confused, doesn't your leaderboard clearly show it is the most accurate model? It's number one in the leaderboard. Am I missing something?
Figured it out. 3.5 Flash and 3.5 Transcribe are different models
Still no real time diarization beyond 3 people (and even then experimental) when others do it very well, like Soniox and Deepgram. For something like meeting notes this is critical. Not sure what the issue is to implement it, maybe that's not Google's use case in mind and rather it's about personal Rambling as the feature on Pixels shows, which uses this model.
> maybe that's not Google's use case in mind
That would be interesting when they also own Google Meet.
Yes, I specifically avoided mentioning that because they don't need to diarize from audio directly as they know exactly which participant is speaking at all times.
Except when you have multiple participants in the same room -- which happens pretty frequently in many business environments.
They have access to each individual’s stream in that case in order to diarize.
Annoying that they don't include pricing info in new models, here it is [1]:
Gemini 3.5 Transcribe Live (Per 1M tokens in USD):
Gemini 3.5 Transcribe: [1] https://ai.google.dev/gemini-api/docs/pricing#gemini-3.5-tra...I am not sure if "Word Error Rate" captures what has always been wrong with transcription. My biggest complaint is that it inserts sentence breaks in random places, then fails to evaluate the result, even though it is obviously wrong. Then I have to go fix it which can be harder than having just typed it myself, due to the difficulty of positioning the Android cursor, the fact that it automatically capitalizes if you delete a capital letter, etc. And much of the time I fail to notice the errors until later.
have another model do a pass to clean it up, saw a demo of local STT where someone did this, can fix a lot of things, especially with gotchas for the STT model in a clean-transcript.md
I haven't tried it, but this looked promising for that exact task: https://huggingface.co/superwhisper/s1-mini
I think the model can even evaluate itself. If it looks afterward at an output like "do you. Want to get lunch?" in the absence of affirmative evidence that the user wanted it that way, it should be able to see that it goofed.
Hopefully YouTube automatic captions improve with this
Very impressive, including the ability to hit fn in any text field and say "generate an image ...".
I'd love to know how this handles proper subtitle formatting. I'm in the process of learning many languages, and being able to cross check my own understanding with film and video would be fantastic.
Where does the compute happen?
I suppose it's a cloud thing?
Yep.
Does somebody knows what top locales list is sampled from? their own usage data? Also when they will use their AI to give better directions in Waze?
Google is crushing it! They’re going toe to toe against the European giant Mistral with all these small useful models and they have a chance of winning! Never have I been more excited about Google’s capabilities.
Is this sarcasm?
Meant as a joke in good spirit. I do appreciate what Google is doing but I am rooting for them to also compete on the frontier.
It can't possibly be anything else.