🎙️ I Built a Local AI Transcriber for a Friend Who Works With Too Much Audio and Video
Description and use
The app automatically transcribes audio and video, synchronizes text with media, and exports in multiple formats, running locally for privacy.
Problem
They had useful information trapped inside audio and video, but getting that information into searchable, readable text was slow and annoying.
Outcome
The application reduces transcription effort, supports 99+ languages, offers multiple Whisper model sizes, and allows exporting in several formats while running locally for privacy.
Workflow
- Upload or record audio/video
- FFmpeg probes media and extracts/normalizes audio to 16 kHz mono WAV
- Whisper performs speech recognition and generates timestamped transcript
- Server-Sent Events stream progress to browser
- User can view synchronized transcript, seek by timestamp, and export in .txt, .srt, .vtt, .json, or .wav formats
Evidence
They had useful information trapped inside audio and video, but getting that information into searchable, readable text was slow and annoying.
The application supports both audio and video files. For video, FFmpeg extracts and normalizes the audio into the 16 kHz, 16-bit mono PCM format expected by Whisper before sending it through the speech-to-text pipeline.
With an open model such as Whisper, the application can be built around local inference instead of requiring every recording to leave the user's machine.
- Record date: 2026-10-03T17:41:11.112Z