Cam SmithAI Studio

← All models

Computational Model Studio · Model release

Parakeet TDT 0.6B v2

Transcribes English speech audio into text with punctuation, capitalization, and timestamp support.

Developer: NVIDIA · Last verified:

What it is designed to do

Convert documented 16 kHz mono English speech audio to a text transcript.

Primary contribution
perception
Input
audio
Output
text (speech transcript, word timestamps)

Documented uses: English speech transcription for captions or voice interfaces.

Access and deployment

  • Parakeet TDT 0.6B v2 model files — available. Public publisher repository; review applicable terms before use. Artifact documentation Local hardware requirements have not been verified. The full artifact-specific rights document has not been separately reviewed; consult the publisher repository conditions.
Technical profile and review notes
Construction
learned
Representation
neural
Normalized architecture tags
No normalized architecture tag assigned
Lifecycle
Unknown from reviewed sources

Inclusion rationale: Adds audio input and speech recognition, which the seed cannot represent.

Limitations

  • Publisher input specification is 16 kHz mono audio; this record does not claim other language coverage or local memory feasibility.

Sources and related Studio records

Sources

  1. Parakeet TDT 0.6B v2 model cardNVIDIA · checked 2026-09-30

Connections