Transcribe Audio Files

Open Transcribe in the sidebar to process existing media without live recording.

Add Files

Drag files into the view, click Choose Files when empty, or Add when a queue is present.

Supported extensions include wav, mp3, m4a, aiff, mp4, mov, aac, flac, caf, amr, ogg, oga, opus, and 3gp. Processing uses the media's audio track.

Choose a Mode

Select an enabled Mode in the top bar before starting. Files use its transcription model, language, paragraph formatting, word replacements, and optional enhancement provider, model, and prompt.

File processing runs batch transcription, not the Mode's live streaming path. Choose a batch-capable model; Cartesia Ink 2 cannot process imported files.

Files do not capture live selected-text, clipboard, or screen context, even if the Mode has those sources enabled. Spoken word triggers do not switch the selected Mode during file processing.

The Mode's delivery options do not run here: files do not paste into the active app, open a Respond conversation, execute Custom Commands, or Auto Send. Results stay in the queue and History.

Process the Queue

Click Start to process pending files sequentially. Cancel stops processing and returns in-progress items to pending. A working model or provider connection is required.

Results and History

Expand a completed row to view its text. Use Original and Enhanced tabs when enhancement produced a result, then copy or save the visible text.

Each completed file also creates a History entry with model and Mode metadata, timings, and a saved converted audio recording. VoiceInk does not modify your source media file.

Clear cancels work and removes queue items from this view. Results already saved in History remain until you delete them or retention cleanup removes them.