Nemotron 3.5 ASR is a multilingual, streaming Automatic Speech Recognition (ASR) model engineered to deliver high-quality multilingual transcription across both low-latency streaming and high-throughput batch workloads. Developed by NVIDIA, this 600M parameter model is a multilingual extension of nvidia/nemotron-speech-streaming-en-0.6b, adding language-ID prompt conditioning to support transcription across 40 language-locales from a single model.
Code, documentation, and weights from the model are publicly available in the official Hugging Face repository.
Nemotron 3.5 ASR transcribes speech into text with native support for punctuation and capitalization. It offers runtime flexibility with configurable chunk sizes, including 80ms, 160ms, 320ms, 560ms, and 1120ms.
By leveraging a state-of-the-art Cache-Aware FastConformer-RNNT architecture, the model eliminates redundant overlapping computations common in traditional "buffered" streaming. This allows it to process only new audio chunks while reusing cached encoder context, significantly improving computational efficiency and minimizing end-to-end delay without sacrificing accuracy.
Key traits of Nemotron 3.5 ASR:
target_lang=auto).The model consists of a cache-aware streaming Parakeet (FastConformer) encoder with an RNN-T decoder and language-ID prompt conditioning. It features 24 encoder layers and maintains caches for all encoder self-attention and convolution layers to eliminate redundant computations.
The language-ID prompt is fused with the acoustic representation as follows:
Nemotron 3.5 ASR is intended for:
Limitations:
To train, fine-tune, or perform inference with this model, you will need to install the NVIDIA NeMo Framework.
import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.ASRModel.from_pretrained(model_name="nvidia/nemotron-3.5-asr-streaming-0.6b")
cd NeMo
python examples/asr/asr_cache_aware_streaming/speech_to_text_cache_aware_streaming_infer.py \
model_path=<model_path> \
dataset_manifest=<dataset_manifest> \
batch_size=<batch_size> \
target_lang=<lang_id> \
att_context_size="[56,13]" \
strip_lang_tags=true \
output_path=<output_folder>