NeMo

This page lists models from NeMo.

Speech Recognition models

• NeMo CTC-based models
  • How to export models from NeMo to sherpa-onnx
    • Step 1: Export model.onnx
    • Step 2: Add metadata
    • Step 3: Obtain model.int8.onnx
    • Step 4: Obtain tokens.txt
  • English
    • sherpa-onnx-nemo-parakeet_tdt_ctc_110m-en-36000-int8 (English, 英语)
      • Download the model
      • Decode wave files
      • Speech recognition from a microphone
      • Speech recognition from a microphone with VAD
    • stt_en_citrinet_512
      • Download the model
      • Decode wave files
    • stt_en_conformer_ctc_small
      • Download the model
      • Decode wave files
    • stt_en_conformer_ctc_medium
      • Download the model
      • Decode wave files
    • stt_en_conformer_ctc_large
      • Download the model
      • Decode wave files
  • Russian
    • sherpa-onnx-nemo-ctc-giga-am-v2-russian-2025-04-19
      • Download the model
      • Decode wave files
      • Speech recognition from a microphone
      • Speech recognition from a microphone with VAD
    • sherpa-onnx-nemo-ctc-giga-am-russian-2024-10-24
      • Download the model
      • Decode wave files
      • Speech recognition from a microphone
      • Speech recognition from a microphone with VAD
  • Japanese
    • sherpa-onnx-nemo-parakeet-tdt_ctc-0.6b-ja-35000-int8 (Japanese, 日语)
      • Download the model
      • Decode wave files
      • Speech recognition from a microphone
      • Speech recognition from a microphone with VAD
• NeMo transducer-based Models
  • sherpa-onnx-nemo-parakeet-unified-en-0.6b-int8-non-streaming (English)
    • Android APK for real-time speech recognition
    • Download the model
    • Decode wave files
    • Real-time/Streaming Speech recognition from a microphone with VAD
    • Speech recognition from a microphone
    • Speech recognition from a microphone with VAD
    • Decode a long audio file with VAD
  • sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8 (25 European Languages)
    • Colab
    • Huggingface space
    • Android APK for real-time speech recognition
    • Download the model
    • Decode wave files
    • Real-time/Streaming Speech recognition from a microphone with VAD
    • Speech recognition from a microphone
    • Speech recognition from a microphone with VAD
    • Decode a long audio file with VAD
  • sherpa-onnx-nemo-parakeet-tdt-0.6b-v2-int8 (English, 英语)
    • Android APK for real-time speech recognition
    • Download the model
    • Decode wave files
    • Real-time/Streaming Speech recognition from a microphone with VAD
    • Speech recognition from a microphone
    • Speech recognition from a microphone with VAD
    • Decode a long audio file with VAD
    • RTF on RK3588 with Cortex A76 CPU
  • sherpa-onnx-nemo-transducer-giga-am-v2-russian-2025-04-19 (Russian, 俄语)
    • Android APK for real-time speech recognition
    • Download the model
    • Decode wave files
    • Real-time/Streaming Speech recognition from a microphone with VAD
    • Speech recognition from a microphone
    • Speech recognition from a microphone with VAD
  • sherpa-onnx-nemo-transducer-giga-am-russian-2024-10-24 (Russian, 俄语)
    • Download the model
    • Decode wave files
    • Real-time/Streaming Speech recognition from a microphone with VAD
    • Speech recognition from a microphone
    • Speech recognition from a microphone with VAD
• Non-streaming Canary models
  • sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8 (English + Spanish + German + French, 英语+西班牙语+德语+法语)
    • Download the model
    • Decode wave files
      • Input English, output English
      • Input English, output German
      • Input English, output Spanish
      • Input English, output French
      • Input German, output English
      • Input German, output German
    • Python API examples
• Nemotron ASR Streaming
  • sherpa-onnx-nemotron-speech-streaming-en-0.6b-560ms-int8-2026-04-25 (English)
    • Export to ONNX
    • Pre-built Android APK for real-time speech recognition
    • Download the model
    • Decode a wave file
    • Real-time speech recognition from a microphone

Speaker Embedding models

Please see Speaker Identification.

Models from NeMo are prefixed with nemo, e.g.,

• nemo_en_titanet_small.onnx

• nemo_en_titanet_large.onnx

• nemo_en_speakerverification_speakernet.onnx

You can find the export scripts at https://github.com/k2-fsa/sherpa-onnx/tree/master/scripts/nemo/speaker-verification