Nemotron Streaming

Nemotron-Speech-Streaming-0.6B is NVIDIA's low-latency English streaming ASR: a cache-aware FastConformer encoder paired with an RNN-T decoder, with native punctuation and capitalization emitted as regular BPE tokens. The CoreML bundle on this site ships with an INT8-palettized encoder and runs on the Apple Neural Engine.

What it is

• Native punctuation and capitalization — no post-processor needed; periods, commas, and case are part of the vocabulary
• 600M parameters — larger than Parakeet-EOU (120M), so transcription quality is materially higher on challenging audio
• Cache-aware FastConformer — 24 encoder layers with attention + conv caches flowing chunk-to-chunk for continuous context
• Four chunk sizes — 80, 160, 560, 1120 ms per inference step (160 ms is the default and the variant published today)
• English only — trained on English speech; for multilingual use see Qwen3-ASR or Omnilingual ASR

Architecture

Three CoreML models pipelined per audio chunk:

No EOU head

Unlike Parakeet-EOU, Nemotron does not emit a dedicated end-of-utterance token. Two ways to segment continuous audio into utterances:

1. External VAD — pair the session with Silero VAD; on sustained silence, call finalize() to commit the current utterance and createSession() for the next one.
2. Punctuation boundary — when the partial transcript ends in ., ?, or !, treat that as a natural commit cue. No extra model, but depends on the audio actually inducing terminal punctuation.

Model

Upstream: nvidia/nemotron-speech-streaming-en-0.6b (NeMo .nemo checkpoint).

Quick start — batch transcription

Conforms to SpeechRecognitionModel, so it drops into any code path that takes a generic STT model:

import NemotronStreamingASR

let model = try await NemotronStreamingASRModel.fromPretrained()
let text = try model.transcribeAudio(audioSamples, sampleRate: 16000)

Quick start — async streaming

for await partial in model.transcribeStream(audio: samples, sampleRate: 16000) {
    if partial.isFinal { print("FINAL: \(partial.text)") }
    else               { print("... \(partial.text)") }
}

Each PartialTranscript carries text, isFinal (true only for the last partial after finalize()), confidence, and a monotonic segmentIndex.

Long-lived session API (mic input)

let session = try model.createSession()

// each mic chunk:
let partials = try session.pushAudio(float32Chunk16kHz)
for p in partials { showPartial(p.text) }   // isFinal is false mid-stream

// when the utterance ends (VAD silence or explicit stop):
let trailing = try session.finalize()
for p in trailing { commit(p.text) }

CLI

audio transcribe recording.wav --engine nemotron                    # batch
audio transcribe recording.wav --engine nemotron --stream           # streaming final
audio transcribe recording.wav --engine nemotron --stream --partial # with partials

Nemotron vs Parakeet-EOU