22h ago

Nick Dobos, prompt engineer at The Browser Company, says mid-conversation model switching forces a reset because thinking tokens cannot translate

Prompt caching cuts costs by up to 80% instead

Sentiment

Pos100%

Neg0%

Positive users highlight custom compression flows enabling seamless model switching in AI chats without double billing or context loss.

1 comment with sentiment.

Nick Dobos, prompt engineer at The Browser Company, says mid-conversation model switching forces a reset because thinking tokens cannot translate · Digg