RadianVector

Making AI & LLMs fast, efficient and safe. Helping teams and companies optimize GPU Inference.

[RadianVector @RadianVector]

If you're curious about how LLM inference works, start here: Prefill vs. Decode.

Includes original measurements and firsthand experience running Qwen3.8-27B. https://t.co/69xEb8B7Fo

[RadianVector @RadianVector]

@AppaloreLabs - guessing you mean this - https://t.co/95iul7NwKR - but looks like that is API only - not open weights yet. Qwen has Qwen-Audio-3.1-ASR-Next - which is the one with diarization - and combines transcription and diarization in one model.

[Qwen @Alibaba_Qwen]

⚡ Meet Qwen-Audio-3.1! ASR, TTS & Realtime are fully upgraded, joined by two new models: TTS-Next for audio creation and ASR-Next for audio understanding.

Five models, one complete audio stack: understanding, generation, interaction & creation.

Plus big price cuts across the lineup: TTS ~70% off, Realtime ~85% off, and ASR up to 95% off.

Highlights: 🥳
- ASR: stronger multilingual & dialect recognition, plus native polishing that auto-removes fillers & repetitions for cleaner, more logical transcripts.
- ASR-Next: supports multi-speaker ASR with speaker labels, timestamps & aligned transcripts, and understands emotions, ambient & machine sounds for sound captioning, event localization, audio QA & reasoning.
- TTS: multilingual & dialect synthesis with natural cross-lingual voice transfer; control emotion, speed & style via simple instructions.
- TTS-Next: unified LM + diffusion framework generating voice, sound effects & background audio in one pass for audiobooks, podcasts, games & ads.
- Realtime: speak & listen at once with anytime interruption, just like a real call; it even slows down and responds empathetically when it senses a low mood.

Unlock the full potential of Qwen-Audio-3.1! 👇
- Blog: https://t.co/e0M8wvj8Kg
- Qwen-Audio-3.1-ASR:
https://t.co/88lPQTPxyz
- Qwen-Audio-3.1-Realtime:
https://t.co/Y63sdtK2AC
- More APIs: coming soon @qwen_cloud

[RadianVector @RadianVector]

For the voice AI enthusiasts - Nvidia launched their new model today that a ton of people are tweeting about - but note that it is a "Diarization" only model - ie - the model tells you who spoke when, not what they said! You need to couple this with a ASR model - an automatic speech recognization model - to get the text.

[Andi Marafioti @andimarafioti]

Voice agents still don’t understand who’s speaking to them. That’s a huge gap compared with humans, hidden by all the “phone-call” demos. But that changes today!

NVIDIA is open-sourcing Nemotron 3 Diarization: a model that can reliably track speakers in live conversations, under a commercial-friendly license! In my tests, the quality is really good with one-second speech chunks. So we can use it for voice agents!

I tested it with Reachy Mini and speech-to-speech running on a DGX Spark. It’s super fun to see the robot notice a new voice, ask for a name, and remember it.

The model has day-zero integration with Transformers!

Kudos to the NVIDIA team for shipping useful tools for the whole community!

[RadianVector @RadianVector]

Nice! We are hoping to try this out soon!

[NVIDIA AI @NVIDIAAI]

When several people talk at once, a transcript can get messy fast.

Our new Nemotron 3 Diarization model tracks who spoke when, even when voices overlap. It handles up to eight speakers, has 100M parameters, and is now available on @huggingface 🤗

[RadianVector @RadianVector]

Would be great to see how Bonsai 2 benchmarks against various quantized Qwen3.8-27B models across these configurations — especially with the PTQ1_0/PQ2_0 + Hadamard setup that is required for the Bonsai 2 model. @RadianVector - yet another thing for you to do 😀 https://t.co/aI2DWGu2u3

[RadianVector @RadianVector]

The upcoming tutorial on AI and LLMs is designed to demystify what happens inside these systems & provide a clearer understanding of how AI works - as it becomes increasingly relevant to all of us

If you have questions, topics you’d like us to explain in more depth, let us know!

[RadianVector @RadianVector]

Learn about the various inference settings and the inference stack you can use on your GPU. https://t.co/tyoQZOBSAb - in night (dark) mode. We will be launching an easy to understand AI, LLM tutorial with great graphics - follow and connect if interested.

[RadianVector @RadianVector]

Learn about the various inference settings and the inference stack you can use on your GPU. https://t.co/tyoQZOBSAb - in night (dark) mode. We will be launching an easy to understand AI, LLM tutorial with great graphics - follow and connect if interested.

[RadianVector @RadianVector]

Our latest benchmark study on Qwen3.8-27B is now published. Reach out for more info, or if you'd like a similar study on another open-weight model. Study: https://t.co/tyoQZOBSAb

[RadianVector @RadianVector]

RadianVector - Your LLM inference configuration partner.

Evaluate and improve how your LLM runs. Reach out!
https://t.co/2gnEZNqUJ6