Releases·Model Release
NVIDIA releases Nemotron 3 Diarization for speaker diarization on batch and streaming audio
NVIDIA released Nemotron 3 Diarization, a model for speaker diarization (identifying who spoke when) supporting both batch and streaming processing, announced via a Hugging Face blog and weights on Hugging Face. It reportedly outperforms many other diarization models and is evaluated with DER (diarization error rate). A YouTube walkthrough demonstrates it running on DGX Spark with NeMo, including full-podcast transcription exportable as text or SRT — relevant to developers building voice agents.
657 engagement1 sourceWed, Sep 23, 2026, 03:36 PM
AI Overview
- NVIDIA releases Nemotron 3 Diarization, a model that handles both batch and streaming speaker diarization and outperforms other models, scored by DER
- The model runs on DGX Spark with NeMo, exports transcripts as text or SRT, and joins NVIDIA's Nemotron Speech family
Related
Products·launched
speech-swift: Nemotron 3 speaker diarization ported to Apple Silicon for local voice agents
speech-swift is a free Apache-2.0 Swift library and CLI that runs NVIDIA's Nemotron 3 speaker diarization locally on Apple Silicon, ported to Core ML INT8 and MLX INT8. Diarization gives voice agents a timeline of which anonymous speaker is active; no LLM is central, though it complements ASR pipelines. It solves the problem of running diarization without cloud services for Mac developers building voice agents.
speech-swift is a free Apache-2.0 Swift library and CLI that runs NVIDIA's Nemotron 3 speaker diarization locally on Apple Silicon, ported to Core ML INT8 and MLX INT8. Diarization gives voice agents a timeline of which anonymous speaker is active; no LLM is central, though it complements ASR pipelines. It solves the problem of running diarization without cloud services for Mac developers building voice agents.
4 engagement1 sourceredditThu, Sep 24, 2026, 09:05 PM
Releases·Model Releasefeatured
Meta Superintelligence Labs releases Muse Voice Transcribe, first real-time audio perception model
Meta Superintelligence Labs launched Muse Voice Transcribe, a real-time streaming ASR model with diarization for 20+ speakers and endpointing. It is an autoregressive multimodal LLM from the Muse Spark family that processes audio in 80ms chunks and decides per chunk whether to keep listening or emit text, trained with RL on word error rate and delay rewards. It is available via the Meta Model API, Meta AI for Mac, and Muse Code.
Meta Superintelligence Labs launched Muse Voice Transcribe, a real-time streaming ASR model with diarization for 20+ speakers and endpointing. It is an autoregressive multimodal LLM from the Muse Spark family that processes audio in 80ms chunks and decides per chunk whether to keep listening or emit text, trained with RL on word error rate and delay rewards. It is available via the Meta Model API, Meta AI for Mac, and Muse Code.
0301.4k engagement3 sourcesxTue, Sep 1, 2026, 05:26 PM
Developments·Paper
Anuvani-Diarization: Zero-Shot Speaker Diarization Across India's 22 Scheduled Languages
Researchers introduce Anuvani-Diarization, a self-supervised end-to-end segmentation model adapted to Indian conversational speech with uncertainty-aware silver supervision. On the new IndicDiarBench covering all 22 scheduled Indian languages, it achieves the lowest diarization error rate among systems the authors ran and matches its parent model on international benchmarks. The study also shows benchmark annotation conventions materially affect DER scores.
Researchers introduce Anuvani-Diarization, a self-supervised end-to-end segmentation model adapted to Indian conversational speech with uncertainty-aware silver supervision. On the new IndicDiarBench covering all 22 scheduled Indian languages, it achieves the lowest diarization error rate among systems the authors ran and matches its parent model on international benchmarks. The study also shows benchmark annotation conventions materially affect DER scores.
0 engagement1 sourcearxivWed, Oct 14, 2026, 03:12 PM
Products·launched
Speaker-labeled transcription with WhisperX on SageMaker AI
A guide/tutorial deployment of WhisperX on AWS SageMaker AI for speaker-labeled, word-level-accurate transcription of spoken audio. WhisperX combines speech-to-text with speaker diarization to fix generic transcription's second-level timestamp drift and missing speaker labels. Serves teams processing contact-center calls, meetings, podcasts, depositions, and broadcast media at scale.
A guide/tutorial deployment of WhisperX on AWS SageMaker AI for speaker-labeled, word-level-accurate transcription of spoken audio. WhisperX combines speech-to-text with speaker diarization to fix generic transcription's second-level timestamp drift and missing speaker labels. Serves teams processing contact-center calls, meetings, podcasts, depositions, and broadcast media at scale.
0 engagement1 sourcerssThu, Sep 24, 2026, 04:20 PM
Releases·Tool Release
cmpunlocker v0.5 adds ECC support and 4 extra SMs on CMP170hx, with a gen3 unlock teased
The open-source cmpunlocker tool released v0.5, adding ECC support and unlocking 4 additional SMs on NVIDIA's CMP 170hx mining card for free. The poster positions it as a cheaper alternative to NVIDIA's DGX Spark, whose 128GB version rose from $4.7k to $7k while a new 64GB version launches at $5k. A Gen3 unlock is reportedly in development.
The open-source cmpunlocker tool released v0.5, adding ECC support and unlocking 4 additional SMs on NVIDIA's CMP 170hx mining card for free. The poster positions it as a cheaper alternative to NVIDIA's DGX Spark, whose 128GB version rose from $4.7k to $7k while a new 64GB version launches at $5k. A Gen3 unlock is reportedly in development.
111 engagement1 sourceredditWed, Oct 7, 2026, 10:55 AM