Audio Transcriber

About the Audio Transcriber

Audio Transcriber turns a recording into clean, editable text — and then exports it as plain TXT, Markdown with timestamps, Word, SRT, VTT, or raw JSON. Drop in an audio file, pick a model, transcribe, edit any lines that need fixing, and download. Click any timestamp to jump the audio player to that exact moment.

Everything happens inside your browser using OpenAI's Whisper models running through transformers.js. Your audio never leaves your device. The first run downloads the open-source speech model (40–250 MB depending on your choice), caches it permanently in the browser, and works fully offline from then on — including for any future recordings.

What you can control

• Wide format support: MP3, WAV, M4A, OGG, WebM, FLAC, AAC, Opus — files up to 200 MB. Long recordings work, they just take longer.
• Model selection: Pick speed vs accuracy.
  • whisper-tiny.en — ~40 MB, English-only, fastest. Great for clean speech.
  • whisper-base.en — ~80 MB, English-only, balanced (default).
  • whisper-small.en — ~250 MB, English-only, best quality for tough audio.
  • whisper-base — ~80 MB, multilingual. Pick this for non-English recordings.
• Audio player: Built-in playback with scrub control so you can verify timestamps as you edit.
• Editable transcript: Every line is editable in place. Click the timestamp to seek the audio, edit the text, or hit × to drop the line.
• Six export formats:
  • TXT — plain prose, no timestamps.
  • Markdown (.md) — readable text with timestamps inline.
  • Word (.docx) — formatted document, no timestamps.
  • SRT — subtitle file for video players and platforms.
  • VTT — web-native subtitle format for HTML5 video.
  • JSON — raw cue data for programmatic use.
• Custom filename: Type your own or let the tool auto-name based on the source file.

Common uses

• Transcribing interviews, podcasts, and panel discussions into editable text.
• Generating searchable notes from lectures, meetings, and webinars.
• Producing subtitle files (SRT/VTT) for videos hosted on YouTube, Vimeo, or self-hosted players.
• Capturing voice memos and journals as searchable Word documents.
• Drafting blog posts or articles from a spoken recording.
• Building knowledge bases from recorded sales calls, customer interviews, or research sessions.
• Pre-transcribing audio before sending it to a human editor for polish — much faster than starting from scratch.

Tips for the best results

• Use the cleanest audio you have. Whisper is robust but noticeably better on recordings where speech is clearly above the noise floor.
• For English-only recordings, the .en models are both faster and noticeably more accurate than the multilingual version.
• Start with whisper-base.en. It's the sweet spot — fast enough for long files, accurate enough for most spoken content. Step up to small.en only when you need it.
• The first run downloads the model. After that, every subsequent transcription is offline and instant to start.
• Click timestamps to seek the audio while editing — fastest way to verify a tricky word or proper noun.
• Use SRT or VTT exports when uploading to YouTube. Use TXT or Word for blog drafts and meeting notes.
• Pre-trim long recordings with Audio Slicer or Audio Merger if you only need part of the audio — transcribing a 10-minute clip is much faster than a 2-hour one.

No accounts, no upload queue, no per-minute charges — drop a file, pick a model, edit the result, and export. Everything stays on your device.

Related tools