[HN Gopher] Qwen3-Omni-Flash-2025-12-01:a next-generation native...
       ___________________________________________________________________
        
       Qwen3-Omni-Flash-2025-12-01:a next-generation native multimodal
       large model
        
       Author : pretext
       Score  : 164 points
       Date   : 2025-12-10 16:13 UTC (6 hours ago)
        
 (HTM) web link (qwen.ai)
 (TXT) w3m dump (qwen.ai)
        
       | dvh wrote:
       | I asked: "How many resistors are used in fuzzhugger phantom
       | octave guitar pedal?". It replied 29 resistors and provided a
       | long list. Answer is 2 resistors:
       | https://tagboardeffects.blogspot.com/2013/04/fuzzhugger-phan...
        
         | iFire wrote:
         | > How many resistors are used in fuzzhugger phantom octave
         | guitar pedal?
         | 
         | Weird, as someone not having a database of the web, I wouldn't
         | be able to calculate either result.
        
           | iFire wrote:
           | I tend to pick things where I think the answer is in the
           | introduction material like exams that test what was taught.
        
           | dvh wrote:
           | "I don't know" would be perfectly reasonable answer
        
             | MaxikCZ wrote:
             | I feel like theres a time in near future where LLMs will be
             | too cautious to answer any questions they arent sure about,
             | and most of the human effort will go into pleading the LLM
             | to at least try to give an answer, which will almost always
             | be correct anyways.
        
               | plufz wrote:
               | That would be a great if you could have a setting like
               | temperature 0.0-1.0 (Only answer if you are 100% to guess
               | as much as you like).
        
               | littlestymaar wrote:
               | It's not going to happen as the user would just leave the
               | platform.
               | 
               | It would be better for most API usage though, as for
               | business doing just a fraction of the job with 100%
               | accuracy is often much preferable than claiming to do
               | 100% but 20% is garbage.
        
           | kaoD wrote:
           | > as someone not having a database of the web, I wouldn't be
           | able to calculate either result
           | 
           | And that's how I know you're not an LLM!
        
         | esafak wrote:
         | This is just trivia. I would not use it to test computers -- or
         | humans.
        
           | parineum wrote:
           | Everything is just trivia until you have a use for the
           | answer.
           | 
           | OP provided a we link with the answer, aren't these models
           | supposed to be trained on all of that data?
        
             | esafak wrote:
             | There is nothing useful you can do with this information.
             | You might as well memorize the phone book.
             | 
             | The model has a certain capacity -- quite limited in this
             | case -- so there is an opportunity cost in learning one
             | thing over another. That's why it is important to train on
             | quality data; things you can build on top of.
        
             | DennisP wrote:
             | Just because it's in the training data doesn't mean the
             | model can remember it. The parameters total 60 gigabytes,
             | there's only so much trivia that can fit in there so it has
             | to do lossy compression.
        
           | littlestymaar wrote:
           | It's good way to assess the model with respect to
           | hallucinations though.
           | 
           | I don't think a model should know the answer, but it must be
           | able to know that it doesn't know if you want to use it
           | reliably.
        
             | esafak wrote:
             | No model is good at this yet. I'd expect the flagships to
             | solve the first.
        
         | brookst wrote:
         | Where did you try it? I don't see this model listed in the
         | linked Qwen chat.
        
         | strangattractor wrote:
         | Maybe it thinks some of those 29 are in series:)
        
       | mettamage wrote:
       | I wonder if with that music analysis mode, you can also make your
       | own synths
        
       | sosodev wrote:
       | Does Qwen3-Omni support real-time conversation like GPT-4o?
       | Looking at their documentation it doesn't seem like it does.
       | 
       | Are there any open weight models that do? Not talking about
       | speech to text -> LLM -> text to speech btw I mean a real voice
       | <-> language model.
       | 
       | edit:
       | 
       | It does support real-time conversation! Has anybody here gotten
       | that to work on local hardware? I'm particularly curious if
       | anybody has run it with a non-nvidia setup.
        
         | dsrtslnd23 wrote:
         | it seems to be able to do native speech-speech
        
           | sosodev wrote:
           | It does for sure. I did some more digging and it does real-
           | time too. That's fascinating.
        
         | red2awn wrote:
         | None of inference frameworks (vLLM/SGLang) supports the full
         | model, let alone non-nvidia.
        
           | sosodev wrote:
           | That's unfortunate but not too surprising. This type of model
           | is very new to the local hosting space.
        
           | AndreSlavescu wrote:
           | We actually deployed working speech to speech inference that
           | builds on top of vLLM as the backbone. The main thing was to
           | support the "Talker" module, which is currently not supported
           | on the qwen3-omni branch for vLLM.
           | 
           | Check it out here:
           | https://models.hathora.dev/model/qwen3-omni
        
             | red2awn wrote:
             | Nice work. Are you working on streaming input/output?
        
               | AndreSlavescu wrote:
               | Yeah, that's something we currently support. Feel free to
               | try the platform out! No cost to you for now, you just
               | need a valid email to sign up on the platform.
        
             | sosodev wrote:
             | Is your work open source?
        
         | ivape wrote:
         | That's exciting. I doubt there are any polished voice chat
         | local apps yet that you can easily plug this into (I doubt the
         | user experience is "there" yet). Even stuff like Silly Tavern
         | is near unusable, lots of work to be done on the local front.
         | Local voice models are what's going to enable that whole
         | Minority Report workflow soon enough (especially if commands
         | and intent are determined at the local level, and the meat of
         | the prompt is handled by a larger remote model).
         | 
         | This is part of programming that I think is the new field.
         | There will be tons of work for those that can build the new
         | workflows which will need to be primarily natural language
         | driven.
        
           | sosodev wrote:
           | I did find this app: https://github.com/gabber-dev/gabber
           | 
           | The creator posted a little demo of it working with Qwen3
           | Omni that is quite impressive:
           | https://www.youtube.com/watch?v=5DBFVe3cLto
           | 
           | He didn't include any details regarding how the model was
           | running though
        
         | potatoman22 wrote:
         | From what I can tell, their official chat site doesn't have a
         | native audio -> audio model yet. I like to test this through
         | homophones (e.g. record and record) and asking it to change its
         | pitch or produce sounds.
        
       | binsquare wrote:
       | Does anyone else find that there's hard to pin down reason of
       | life-lessness in the speech of these voice models?
       | 
       | Especially in the fruit pricing portion of the video for this
       | model. Sounds completely normal but I can immediately tell it is
       | ai. Maybe it's intonation or the overly stable rate of speech?
        
         | colechristensen wrote:
         | I'm perfectly ok with and would prefer an AI "accent".
        
         | esafak wrote:
         | > Sounds completely normal but I can immediately tell it is ai.
         | 
         | Maybe that's a good thing?
        
         | sosodev wrote:
         | I think it's because they've crammed vision, audio, multiple
         | voices, prosody control, multiple languages, etc into just 30
         | billion parameters.
         | 
         | I think ChatGPT has the most lifelike speech with their voice
         | models. They seem to have invested heavily in that area while
         | other labs focused elsewhere.
        
         | Lapel2742 wrote:
         | IMHO it's not lifeless. It's just not overly emotional. I
         | definitely prefer it that way. I do not want the AI to be
         | excited. It feels so contrived.
         | 
         | On the video itself: Interesting, but "ideal" was pronounced
         | wrong in German. For a promotional video, they should have
         | checked that with native speakers. On the other hand its at
         | least honest.
        
           | nunodonato wrote:
           | I hate with a passion the over-americanized "accent" of
           | chatgpt voices. Give me a bland one any day of the week
        
         | vessenes wrote:
         | I'm not convinced its end-to-end multimodal - in that case,
         | you'll have a speech synthesis section and this will be some of
         | the result. You could test by having it sing or do some
         | accents, or have it talk back to you in an accent you give it.
        
       | rarisma wrote:
       | GPT4o in the charts is crazy.
        
         | BoorishBears wrote:
         | Why? gpt-realtime is finalized gpt-4o. Gemini Live is still
         | 2.5.
         | 
         | Not their fault frontier labs are letting their speech to
         | speech offerings languish.
        
       | banjoe wrote:
       | Wow, crushing 2.5 Flash on every benchmark is huge. Time to move
       | all of my LLM workloads to a local GPU rig.
        
         | embedding-shape wrote:
         | Just remember to benchmark it yourself first with you private
         | task collection, so you can actually measure them against each
         | other. Pretty much any public benchmark is unreliable at this
         | moment, and making model choices based on other's benchmarks is
         | bound to leave you disappointed.
        
           | MaxikCZ wrote:
           | This. Last benchmarks of DSv3.2spe hinted at beating
           | basically everything, yet in my testing even sonnet is miles
           | ahead both in terms of speed and accuracy
        
         | red2awn wrote:
         | Why would you use an Omni model for text only workload... There
         | is Qwen3-30B-A3B.
        
       | gardnr wrote:
       | This is a 30B parameter MoE with 3B active parameters and is the
       | successor to their previous 7B omni model. [1]
       | 
       | You can expect this model to have similar performance to the non-
       | omni version. [2]
       | 
       | There aren't many open-weights omni models so I consider this a
       | big deal. I would use this model to replace the keyboard and
       | monitor in an application while doing the heavy lifting with
       | other tech behind the scenes. There is also a reasoning version,
       | which might be a bit amusing in an interactive voice chat if it
       | pronounces the thinking tokens while working through to a final
       | answer.
       | 
       | 1. https://huggingface.co/Qwen/Qwen2.5-Omni-7B
       | 
       | 2. https://artificialanalysis.ai/models/qwen3-30b-a3b-instruct
        
         | gardnr wrote:
         | I can't find the weights for this new version anywhere. I
         | checked modelscope and huggingface. It looks like they may have
         | extended the context window to 200K+ tokens but I can't find
         | the actual weights.
        
           | pythux wrote:
           | They link to: https://huggingface.co/collections/Qwen/qwen3-o
           | mni-68d100a86... from the blog post but it does seem like
           | this redirects to their main space on HF so maybe they didn't
           | yet make the model public?
        
         | olafura wrote:
         | Looks like it's not open source:
         | https://www.alibabacloud.com/help/en/model-studio/qwen-omni#...
        
           | coder543 wrote:
           | No... that website is not helpful. If you take it at face
           | value, it is claiming that the previous Qwen3-Omni-Flash
           | wasn't open either, but that seems wrong? It is very common
           | for these blog posts to get published before the model
           | weights are uploaded.
        
             | red2awn wrote:
             | The previous -Flash weight is closed source. They do have
             | weights for the original model that is slightly behind in
             | performance
             | https://huggingface.co/Qwen/Qwen3-Omni-30B-A3B-Instruct
        
               | coder543 wrote:
               | Based on things I had read over the past several months,
               | Qwen3-Flash seemed to just be a weird marketing term for
               | the Qwen3-Omni-30B-A3B series, not a different model. If
               | they are not the same, then that is
               | interesting/confusing.
        
               | red2awn wrote:
               | It is an in-house closed weight model for their own chat
               | platform, mentioned in Section 5 of the original paper:
               | https://arxiv.org/pdf/2509.17765
               | 
               | I've seen it in their online materials too but can't seem
               | to find it now.
        
         | tensegrist wrote:
         | > There is also a reasoning version, which might be a bit
         | amusing in an interactive voice chat if it pronounces the
         | thinking tokens while working through to a final answer.
         | 
         | last i checked (months ago) claude used to do this
        
         | andy_xor_andrew wrote:
         | > This is a 30B parameter MoE with 3B active parameters
         | 
         | Where are you finding that info? Not saying you're wrong; just
         | saying that I didn't see that specified anywhere in the linked
         | page, or on their HF.
        
           | plipt wrote:
           | The link[1] at the top of their article to HuggingFace goes
           | to some models named Qwen3-Omni-30B-A3B that were last
           | updated in September. None of them have "Flash" in the name.
           | 
           | The benchmark table shows this Flash model beating their
           | Qwen3-235B-A22B. I dont see how that is possible if it is a
           | 30B-A3B model.
           | 
           | I don't see a mention of a parameter count anywhere in the
           | article. Do you? This may not be an open weights model.
           | 
           | This article feels a bit deceptive
           | 
           | 1: https://huggingface.co/collections/Qwen/qwen3-omni
        
         | red2awn wrote:
         | This is a stack of models:
         | 
         | - 650M Audio Encoder
         | 
         | - 540M Vision Encoder
         | 
         | - 30B-A3B LLM
         | 
         | - 3B-A0.3B Audio LLM
         | 
         | - 80M Transformer/200M ConvNet audio token to waveform
         | 
         | This is a closed source weight update to their Qwen3-Omni
         | model. They had a previous open weight release
         | Qwen/Qwen3-Omni-30B-A3B-Instruct and a closed version
         | Qwen3-Omni-Flash.
         | 
         | You basically can't use this model right now since none of the
         | open source inference framework have the model fully
         | implemented. It works on transformers but it's extremely slow.
        
         | plipt wrote:
         | I dont think the Flash model discussed in the article is 30B
         | 
         | Their benchmark table shows it beating Qwen3-235B-A22B
         | 
         | Does "Flash" in the name of a Qwen model indicate a model-as-a-
         | service and not open weights?
        
           | red2awn wrote:
           | Flash is a closed weight version of
           | https://huggingface.co/Qwen/Qwen3-Omni-30B-A3B-Instruct (it
           | is 30B but with addtional training on top of the open weight
           | release). They deploy the flash version on Qwen's own chat.
        
             | plipt wrote:
             | Thanks
             | 
             | Was it being closed weight obvious to you from the article?
             | Trying to understand why I was confused. Had not seen the
             | "Flash" designation before
             | 
             | Also 30B models can beat a semi-recent 235B with just some
             | additional training?
        
               | red2awn wrote:
               | They had a Flash variant released alongside the original
               | open weight release. It is also mentioned in Section 5 of
               | the paper: https://arxiv.org/pdf/2509.17765
               | 
               | For the evals it's probably just trained on a lot of the
               | benchmark adjacent datasets compared to the 235B model.
               | Similar thing happened on other model today:
               | https://x.com/NousResearch/status/1998536543565127968 (a
               | 30B model trained specifically to do well in maths get
               | near SOTA scores)
        
       | Aissen wrote:
       | Is this a new proprietary model?
        
       | sim04ful wrote:
       | The main issue I'm facing with realtime responses (speech output)
       | is how to separate non-diegetic outputs (e.g thinking, structured
       | outputs) from outputs meant to be heard by the end user.
       | 
       | I'm curious how anyone has solved this
        
         | artur44 wrote:
         | A simple way is to split the model's output stream before TTS.
         | Reasoning/structured tokens go into one bucket, actual user-
         | facing text into another. Only the second bucket is
         | synthesized. Most thinking out loud issues come from feeding
         | the whole stream directly into audio.
        
           | pugio wrote:
           | There is no TTS here. It's a native audio output model which
           | outputs audio tokens directly. (At least, that's how the
           | other real-time models work. Maybe I've misunderstood the
           | Qwen-Omni architecture.)
        
             | artur44 wrote:
             | True, but even with native audio-token models you still
             | need to split the model's output channels.
             | Reasoning/internal tokens shouldn't go into the audio
             | stream only user-facing content should be emitted as audio.
             | The principle is the same, whether the last step is TTS or
             | audio token generation.
        
       | stevenhuang wrote:
       | Wayback for those that can't reach
       | https://web.archive.org/web/20251210164048/https://qwen.ai/b...
        
       | terhechte wrote:
       | Is there a way to run these Omni models on a Macbook quantized
       | via GGUF or MLX? I know I can run it in LMStudio or Llama.cpp but
       | they don't have streaming microphone support or streaming webcam
       | support.
       | 
       | Qwen usually provides example code in Python that requires Cuda
       | and a non-quantized model. I wonder if there is by now a good
       | open source project to support this use case?
        
         | mobilio wrote:
         | Yes - there is a way: https://github.com/ggml-org/whisper.cpp
        
           | novaray wrote:
           | Whisper and Qwen Omni models have completely different
           | architectures as far as I know
        
         | tgtweak wrote:
         | You can probably follow the vLLM instructions for omni here,
         | then use the included voice demo html to interface with it:
         | 
         | https://github.com/QwenLM/Qwen3-Omni#vllm-usage
         | 
         | https://github.com/QwenLM/Qwen3-Omni?tab=readme-ov-file#laun...
        
       | aschobel wrote:
       | Looks to be API only. Bummer.
        
         | readyplayeremma wrote:
         | The models are right here, one of the first links in the post:
         | https://huggingface.co/collections/Qwen/qwen3-omni
         | 
         | edit: Nevermind, in spite of them linking it at the top, they
         | are the old models. Also, the HF demo is calling their API and
         | not using HF for compute.
        
           | aschobel wrote:
           | It is super confusing. I also thought this initially was open
           | weights.
        
       | devinprater wrote:
       | Wow, just 32B? This could almost run on a good device with 64 GB
       | RAM. Once it gets to Ollama I'll have to see just what I can get
       | out of this.
        
         | plipt wrote:
         | I see that their HuggingFace link goes to some
         | Qwen3-Omni-30B-A3B models that show a last updated date of
         | September
         | 
         | The benchmark table in their article shows Qwen3-Omni-
         | Flash-2025-12-01 (and the previous Flash) as beating
         | Qwen3-235B-A22B. How is that possible if this is only a 30B-A3B
         | model? Also confusing how that comparison column starts out
         | with one model but changes them as you descend down the table.
         | 
         | I don't see any FLASH variant listed on their Hugginface. Am i
         | just missing it or are these specifying a model only used for
         | their API service and there are no open weights to download?
        
         | apexalpha wrote:
         | I run these on a 48gb Mac because of the universal ram.
        
       | vessenes wrote:
       | Interesting - when I asked the omni model at qwen.com what
       | version it was, I got a testy "I don't have a version" and then
       | was told my chat was blocked for inappropriate content. A second
       | try asking for knowledge cutoff got me the more equivocal "2024,
       | but I know stuff after that date, too".
       | 
       | No idea how to check if this is actually deployed on qwen.com
       | right now.
        
         | zamadatix wrote:
         | > No idea how to check if this is actually deployed on qwen.com
         | right now.
         | 
         | Assuming you mean qwen.ai, when you run a query it should take
         | you to chat.qwen.ai with the list of models in the top left.
         | None of the options appear to be the -Omni variant (at least
         | when anonymously accessing it).
        
           | vessenes wrote:
           | Thanks - yes - I did. The blog post suggests clicking the
           | 'voice' icon on the bottom right - that's what I did.
        
         | mh- wrote:
         | For what it's worth, that's not a reliable way to check what
         | model you're interacting with.
        
       ___________________________________________________________________
       (page generated 2025-12-10 23:00 UTC)