How to transcribe audio and video files to text with one API call (MP3, MP4, Google Drive, Dropbox)

Описание и применение

Автоматизированный рецепт транскрипции аудио и видео файлов через один API‑запрос, объединяющий извлечение аудио, разбиение больших файлов, обработку ссылок и возврат текста и SRT.

Задача

Transcribing large or remote audio/video files is cumbersome due to size limits, format requirements, and share‑link handling.

Результат

A 50‑minute MP3 from Dropbox was transcribed in ~40 s for $0.15 (≈$0.003 per minute).

Рабочий процесс

• Provide file URLs (direct or share links) or upload a file
• API extracts audio from video if needed
• Splits large files into 10‑minute chunks
• Transcribes each chunk with Whisper large‑v3
• Adjusts timestamps for each chunk
• Combines text and segments
• Returns JSON with status, fileName, language, duration, text, segments, and optional SRT

Доказательства

A 50-minute MP3 shared through Dropbox was transcribed in about 40 seconds and cost $0.15.

Источник

Whisper wants audio. Pull the audio track out with ffmpeg first, and shrink it while you are there, since speech models work on 16 kHz mono anyway:

Источник

A Google Drive or Dropbox share link opens a preview page, not the file. You need the direct-download form:

Источник