Authors and ModelsSonilo

Sonilo Proofread

Sonilo Proofread
Sonilo · Transcription
POST /v1/audio/transcriptions

Transcribes a video and translates the script into 24 languages, returning editable subtitle tracks you can correct before dubbing.

At a glance

FieldValue
Model idsonilo-proofread
Model release date-
Input modalitiesVideo
Output modalitiesText
Context window-
Weight precision-
Featurestranscription, translation, multilingual, srt_vtt, timestamps
Native inferenceNo
NewYes
Supported endpointsPOST /v1/audio/transcriptions
Alternate model idssonilo/sonilo-proofread

Pricing

ChargeSpecRate
Proofreadper second per language$0.002

Example request

curl https://api.empiriolabs.ai/v1/audio/transcriptions \
  -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \
  -F model=sonilo-proofread \
  -F file=@meeting.mp3

Parameters

ParameterTypeRequiredDefaultDescription
languagesstringno-JSON array of language codes to translate the transcript into. Billed per language. Omit it for the source transcript only. Supported: en, zh_cn, ja, ko, pt, pt_br, es, es_419, de, fr, it, ru, th, ar, tr, vi, id, ta, ml, kn, gu, pa_in, sd_in, hi.
source_languagestringno-The language actually spoken in the video, which helps on short, noisy or mixed-language audio. Omit it to detect the language. Supported: en, zh_cn, ja, ko, pt, pt_br, es, es_419, de, fr, it, ru, th, ar, tr, vi, id, ta, ml, kn, gu, pa_in, sd_in, hi.
video_urlstringno-Source video, as a publicly reachable URL. Up to 300MB and 5 minutes, and it must carry an audio track.

Notes

Languages

en, zh_cn, ja, ko, pt, pt_br, es, es_419, de, fr, it, ru, th, ar, tr, vi, id, ta, ml, kn, gu, pa_in, sd_in, hi

Controls

Supports source video, a list of target languages to translate the transcript into, and an optional expected source language for short or noisy audio.

Limits

  • Source video up to 300MB and 5 minutes
  • The video must carry an audio track
  • Subtitle tracks are returned inline as text, one per language, always including the source language

Billing

  • Charged per second of source video multiplied by the number of target languages, with a 10 second minimum. A transcript-only request counts as one.
  • A request rejected before generation is not billed.

Machine-readable schema: GET https://api.empiriolabs.ai/v1/models/sonilo-proofread.