I’d really really love if we get very light weight smart reasoning models, for example Gemini 3.1 Flash Lite , I recently migrated from Vapi to retellai, and I absolutely love everything, the only things that make this product a 9/10 and not 10/10 are:
-
Not enough choices of models natively within retell
-
Not enough interruption handling settings (ignore this for now)
Me personally I like single prompt agents, these need huge context windows, especially for long calls, so I had to go with Gemini 3 flash, but i tried 2.5 flash lite, way cheaper WAY faster, i loved it but its not smart enough to handle my kind of calls, so i’d like to get access to 3.1 flash lite natively, and also have a more organized model choice drop down, maybe filter by provider like vapi.
Gemini 3.1 flash lite would be a game changer for me, I’m planning to handle 4000/mins a month the cost savings would be great and the customer experience would be way better.