zlacker

[parent] [thread] 6 comments
1. observ+(OP)[view] [source] 2026-02-04 15:53:39
Native diarization, this looks exciting. edit: or not, no diarization in real-time.

https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-26...

~9GB model.

replies(1): >>coder5+Y4
◧
2. coder5+Y4[view] [source] 2026-02-04 16:16:09
>>observ+(OP)
The diarization is on Voxtral Mini Transcribe V2, not Voxtral Mini 4B.
replies(2): >>observ+G8 >>sbroth+de
◧◩
3. observ+G8[view] [source] [discussion] 2026-02-04 16:30:57
>>coder5+Y4
Ahh, yeah, and it's explicitly not working for realtime streams. Good catch!
◧◩
4. sbroth+de[view] [source] [discussion] 2026-02-04 16:54:40
>>coder5+Y4
Do you have experience with that model for diarization? Does it feel accurate, and what's its realtime factor on a typical GPU? Diarization has been the biggest thorn in my side for a long time..
replies(2): >>coder5+ak >>ashenk+RI
◧◩◪
5. coder5+ak[view] [source] [discussion] 2026-02-04 17:22:27
>>sbroth+de
> Do you have experience with that model

No, I just heard about it this morning.

◧◩◪
6. ashenk+RI[view] [source] [discussion] 2026-02-04 19:03:41
>>sbroth+de
You can test it yourself for free on https://console.mistral.ai/build/audio/speech-to-text I tried it on an english-speaking podcast episode, and apart from identying one host as two different speakers (but only once for a few sentences at the start), the rest was flawless from what I could see
replies(1): >>sbroth+Xm1
◧◩◪◨
7. sbroth+Xm1[view] [source] [discussion] 2026-02-04 22:16:03
>>ashenk+RI
Amazing. Thank you.
[go to top]