Speech Recognition Cloud (SRC) is a Windows 10/11 speech recognition application that converts speech to text using cloud processing. It types dictated text directly at the active cursor in most Windows applications (for example Microsoft Word, Outlook, Teams, Chrome, and web forms) rather than using a separate dictation box. Dictation can be started and stopped via configurable hotkeys, hold-to-talk or toggle modes, and optional mouse button activation. Output supports automatic punctuation and spoken punctuation/formatting commands, including new lines/paragraphs and list formatting.
SRC includes workflow controls such as cancelling a dictation before insertion, re-pasting the last dictation via hotkey, optional auto-spacing between dictations, and an option to send messages by automatically pressing Enter after dictation. Users can add custom vocabulary terms and configure text replacements/expansions; higher tiers support advanced replacement rules and reusable templates. Plan-dependent features include AI-assisted modes (e.g., drafting/rewriting/replying) and an optional screen-context feature that uses on-screen text to improve recognition accuracy. The product supports 50+ languages (plan-dependent), while the Medical edition is English-only and includes medical vocabulary support with restricted AI features. SRC requires an internet connection for cloud recognition.