ESP32 Forum

I bought 2 of the esp32-korvo boards with the understanding that I would be able to make voice assistants out of them. But the English recognition examples only target the s3 chips. Is Multinet not compatible with the esp32 in the korvo 1.1? Or is it just that the English example was developed on the s3 and wasn't backported?

ESP32-S3 has AI instructions and faster PSRAM. In general， the performance of ESP32-S3 on AI is 5~10 times that of ESP32.
The parameters of new English Multinet model have exceeded 2M. It is too big to run in real time on ESP32.
We strongly recommend that you use the ESP32-S3 board to run AI-related tasks.

It's very disappointing that I now have $100 worth of useless hardware. How is it that the Chinese version is so much smaller? I can't find the S3 korvo boards in stock anywhere.

For Chinese multinet, we also recommend using ESP32-S3, the performance of ESP32-S3 Chinese multinet has a great improvement.

Looking for a solution to make use of the hardware I have, I'm wondering if there are any examples for doing just wake word on the esp32, then offloading the voice recognition to a separate server.