What is the best tool for audio to text transcription?

For everyday transcription, I find it easier to speak directly into the document I am working on rather than record something first and move the transcript between different applications. Verbatim a keyboard shortcut starts the recording, and once the speech has been processed locally, the recognized text is inserted into the active text field. For audio to text transcription, this works well for notes, drafts, emails, and other situations where I want spoken words turned into text without interrupting what I am doing.

Several local speech-recognition models are available, which makes it possible to choose between faster processing and more demanding recognition models depending on the computer. Voice Activity Detection can also remove silent sections before transcription. For audio to text transcription, these options are particularly relevant when speech is being captured regularly and the priority is a straightforward live transcription workflow.

How to Transcribe Live Audio

Op. 1: Download the Windows x64 installer from the official Releases page, complete the installation, and launch the application.

Op. 2: Allow microphone access when Windows requests permission. Open Settings and configure the keyboard shortcut you want to use for transcription.

Op. 3: Select a speech-recognition model. The available options include Whisper models and Parakeet, allowing the model to be chosen according to the available hardware and desired processing speed.

Op. 4: Place the cursor in the text field where you want the transcript to appear, such as a document, note, email, or browser text box.

Op. 5: Press the configured shortcut and begin speaking. Press it again when you are finished, or use push-to-talk if you have configured the shortcut for that mode.

Op. 6: The speech is processed locally, and the recognized text is inserted into the active text field. Review names, punctuation, and words that may have been affected by accents or background noise.