Your Audio Stays Local
Your audio files never leave your machine. When you interact with Reagent, only text-based commands and metadata (such as file names, track names, and item properties) are sent to process your requests. Audio analysis—including peak detection, silence detection, spectral analysis, UCS Classification and dialogue transcription happens entirely on your local machine.AI Model Provider
Reagent is powered by Anthropic’s Claude models. All text-based processing is handled through Anthropic’s API.Your Data Is Not Used for Training
By default, Anthropic does not use inputs or outputs from their API to train models. Your conversations and commands remain private. For full details, see Anthropic’s official policy: Is my data used for model training?What Data Is Sent
When you send a message in Reagent, the following may be included in the request to Anthropic’s API:- Your text prompt
- Relevant context about your REAPER session (track names, item names, project structure)
- Tool call results (text-based metadata from REAPER operations such as Media Item GUID)