tapWhisper — NVIDIA Canary ONNX

Specifications

Size ~154 MB (INT8 ONNX)

Architecture Conformer

Latency Medium

Language English, Spanish, French

Developer / Creator

NVIDIA (NeMo team), Sherpa ONNX community

License

CC BY 4.0 model; Apache-2.0 Sherpa ONNX runtime

Download Source

Verified Repository Source

Hugging Face Hub / Sherpa ONNX model catalog

NVIDIA Canary 180M Flash

Exact runtime artifacts

sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8.tar.bz2

Model Overview

NVIDIA's Canary is a multilingual speech recognition model. In tapWhisper it transcribes English, Spanish, and French entirely on your Mac through Sherpa ONNX. Transcripts stay in the language you speak — speech translation is not enabled.

Available Model Variants

Model Name	File Size	RAM Usage	Format/Quant	Languages	Description
NVIDIA Canary	154 MB	650 MB	INT8 (ONNX)	EN, ES, FR	NVIDIA Canary 180M Flash. On-device speech recognition in English, Spanish, and French.

Back to tapWhisper