Select the model you would like to use, click the "Start" button and start speaking
For the most accurate translation select Whisper Base. For speed, or for slower systems choose Whisper Tiny Quantized.
Whisper model:
Quantized models:
Quantized models:
Status: not started
[The transcribed text will be displayed here]
Debug output:
Test this link Mic Image Troubleshooting
Your browser needs to support WASM Fixed-width SIMD
The page does some heavy computations. For best results:
- To use a modern web browser (e.g. Chrome, Firefox)
- To use a fast desktop or laptop computer