WhisperARC documentation

What WhisperARC does

WhisperARC runs Whisper on Windows without Python or a terminal. Use it for dictation, system audio, media, translation and local history.

Choose Local, Cloud or OpenAI KEY in Model. Local use is free, Pro and Annual include 9,000 Cloud credits per month, and every plan supports your own OpenAI key with separate OpenAI billing.

Current capabilities

• Dictate with one configurable shortcut in hold-to-talk or start/stop mode, then insert text into the focused editable field.
• Capture a selected Windows output such as speakers or headphones.
• Queue several audio or video files and let WhisperARC work through them one at a time.
• Export text as TXT, SRT, VTT, JSON or DOCX. You can also export audio saved in History in seven formats.
• Create a General glossary and separate project glossaries for names and specialist vocabulary.
• Choose between six local Whisper transcription models and run them on the CPU or a compatible GPU.
• Edit Mail, Message and Code modes or create custom dictation modes linked to Windows apps. The selected app's instructions rewrite microphone results automatically.
• Replay saved audio with transcript highlighting and adjustable speed. Menu shows average dictation speed, dictated words, estimated time saved and transcribed files.

Recommended path

Install WhisperARC and check the selected microphone and model. Then try a short dictation. This is the quickest way to catch a device or shortcut problem before starting a long recording.