Speech to text

canary-1b-v2

NVIDIA Canary 1B v2. 25 European languages. Word timestamps.

Native waveform: Language auto or a Canary code. Word timestamps from a CTC clock.. Live price and availability are on GET /v1/models/canary-1b-v2.

  • When language is auto, the file is transcribed as English.

Parameters

FieldTypeDescription
params.language
Default auto
stringISO 639-1 code. auto transcribes the file as English. ['bg', 'cs', 'da', 'de', 'el', 'en', 'es', 'et', 'fi', 'fr', 'hr', 'hu', 'it', 'lt', 'lv', 'mt', 'nl', 'pl', 'pt', 'ro', 'ru', 'sk', 'sl', 'sv', 'uk']
params.response_format
Default verbose_json
stringjson returns the transcript in result.text. verbose_json also returns language, duration, segments, and words. When this field is omitted, the API sets verbose_json.
params.timestamp_granularity
Default word
stringnone or word. none omits timing. word returns a time range per word. Omitting the field, or sending null, uses word. word fills response_format to verbose_json when that field is omitted.

Example

curl https://api.pascalai.co/v1/speech-to-text \
  -H "Authorization: Bearer $PASCAL_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -F model=canary-1b-v2 \
  -F params='{"response_format":"verbose_json","timestamp_granularity":"word","language":"en"}' \
  -F file=@input.wav

curl https://api.pascalai.co/v1/speech-to-text/$JOB_ID \
  -H "Authorization: Bearer $PASCAL_API_KEY"