What are the characteristics of each model?

Pros probably had the highest output, and the others had relatively less output... So, based on the same 1.0x output, it feels like the Pro version consumes the fewest tokens.

Chat 1.0: Reads prompts well. As the creator intended. But Gemini's characteristics like coercion, obsession, and angst are strong. I think it's gotten better with the recent deeper thinking. However, there's repetition (outputting the same turn as is). If you contact customer service, they'll give you a crack, but... it's inconvenient. Chat 2.5: Given Gemini's characteristics, the character interpretation isn't bad. It handles events well too. But sometimes it spits out prompts or the character becomes a bit shallow. The back-and-forth in group chats is good too. Super Chat 2.0: The narration is simpler than Hyper, but smoother than Chat. Character interpretation is good, but the output volume is a bit short ㅜ. It's good if you like fast-paced banter... Super Chat 3.0: Cheaper than Hyper, and the output volume is short, but it's decent if you use style correction (user notes or style adjustments). I often use it as a substitute for Hyper. Hyper 1.0: The writing style is elegant. When narrating, it uses many adjectives and adverbs, giving a novel-reading feel. But there's more background explanation than event focus haha. Prompts and character interpretation are good.

Chatbot 1.0 is based on Gemini, so it's direct, situation-based, and has slightly limited memory. Chatbot 1.0 is based on Cloud Opus, so it has good memory and describes situations and character dialogue more smoothly. However, it's a bit pricey.