NVIDIA · Transcription model

Parakeet TDT 0.6B v3

nvidia/parakeet-tdt-0.6b-v3

Transcription

Released May 27, 2026

Parakeet TDT 0.6B v3 is NVIDIA's 600M-parameter multilingual speech-to-text model built on the FastConformer-TDT architecture. Trained on the Granary dataset (670,000+ hours of audio), it supports automatic language detection across...

Specifications

Input: wav, mp3, flac, m4a, ogg, webm, aac, opus · base64 JSON, multipart file or URL · up to 25 MB
Output: text; verbose_json adds language, duration, segments and word timestamps
Billing: per second of audio

Use it

POST /api/v1/audio/transcriptions · MCP transcribe · full reference

curl https://omnirail.org/api/v1/audio/transcriptions \
  -H "Authorization: Bearer $OMNIRAIL_KEY" \
  -F model=nvidia/parakeet-tdt-0.6b-v3 \
  -F file=@meeting.mp3