Since 18th January this year I have moved on to recommending students use Gemini to get feedback on their speaking. You can see all my blog posts labelled with Gemini here:
https://onewaytotefl.blogspot.com/search/label/%23Gemini
Sometime in the near future I could adapt the ideas in these blog posts into a submission to the magazine.

Great point!

Cool! You could consider adapting your blogspot post into a submission to the magazine!

The idea is that students record themselves doing a speaking task in class or out. They can then upload their audio (or video) file to Turboscribe and use the Custom ChatGPT connection to ask for feedback.
This is the prompt students can copy and paste in to get feedback on their speaking, but not sadly on their pronunciation [See end of post]:
https://docs.google.com/document/d/1RftjZqIBkxb1CMln3AFuKGypo9AZ2IyZY0KMADZGtR8/edit?usp=sharing

This produces different versions of the transcript for students to learn from. In the notes it says, “If you are using the CEFR scales you may prefer to ask for the next versions to be half a level higher than the previous version. Or one level higher if you want a more dramatic progression.

If you are using the GSE scales you may want to ask for the next versions to be 5 points higher than the previous version. Or 10 points higher if you want a more dramatic progression.

If you are using the IELTS scales you may want to ask for the next versions to be 0.5. points higher than the previous version. Or 1 point higher if you want a more dramatic progression.”

More recently, I have been experimenting with uploading audio (or video) files directly to Gemini as I heard from Frank Andrade that Gemini and Gemini alone can actually listen to the sounds you make pronouncing what you say

Большое спасибо!

Edo Forsythe