Ermine

Ermine.ai introduces an AI tool enabling direct audio transcription from a device's microphone via local processing, avoiding external servers or internet connections. Available on GitHub, users can download audio and transcripts. Initialization of the transcription model on the user's browser precedes transcription, taking a few minutes for the initial 50mb model download and caching. The tool exclusively supports English transcription and may seek microphone access permission. Ermine.ai prioritizes privacy and data security by keeping audio data on the user's device, offering a secure solution. Additionally, the option to save both audio and transcript enhances usability and accessibility for users.

‍