To reasonably control the cost of OpenAI API usage, the following strategies are recommended:1) Setting usage limits: Set a monthly spending cap for the API in the backend of your OpenAI account; 2) Intelligent Mode Switching: local processing (tiny/base model) is used on a daily basis, with temporary switching to cloud mode via the control panel only when high accuracy is required; 3) Audio pre-processing: For recordingCommand+Shift+SpaceShortcut to perform local mute detection first to avoid uploading invalid audio clips; 4) History Management: Clean up the transcription history in the control panel regularly to avoid the SQLite database getting too big and affecting performance; 5) Proxy Naming Filter: Reduce unnecessary API calls by setting the agent name (e.g. "Jarvis") to distinguish between AI commands and regular dictation.
This answer comes from the articleOpenWispr: Privacy-First Speech-to-Text Desktop ApplicationThe
English 
简体中文 
日本語 
Deutsch 
Português do Brasil