MFA Press Conference Transcriber

Configuration
New workflow: The server will fetch and split the audio channels for you. After submitting, you will be able to preview and select a channel before transcription. No download or upload is required on your device.

Switch to the tab that’s playing the audio you’d like to capture, then return here and click Start Recording.

-- dB
Model in use: gemini-2.5-flash-preview-05-20
Controls output randomness. Higher values lead to more "creative" output. Typical values: 0.0-1.0.
Controls output diversity. Lower values limit the scope of responses. Typical values: 0.0-1.0. 0.95 is Google's default.
Optional instructions for the model.