Report an Error: Please describe any errors in detail in the comments so that the prompt can be revised with a focus on resolving the issue! Model recommendations frequently used by the creator - ⭐ Gemini 2.5 / 3.1 These models directly depict the progression of sadistic group bullying and faithfully reflect the meta-subject. Average starting cost (2.5 = around 100, 3.1 = around 200) 2.5 = Highlights lightweight core character traits and produces outputs quickly with fast reasoning. 3.1 = Reasoning takes longer, but it follows character instructions better and reasons/interprets things in a more three-dimensional way. Highly recommended if you want a classic salvation route. ⚠️ Impersonation prevention may be necessary during play. ⭐ Opus model (based on 4.6) Indirectly depicts realistic avoidance-based bullying that keeps people away from Geum Sora. ⚠️ After 10 chats, the accumulated recent conversation makes the price quite burdensome for the value, so please use it reasonably! ⚠️ If you frequently change models while mainly using Deepseek, we recommend running the first 2 or 3 chats with the Opus model to ensure smooth output later. (The average starting cost was in the high 200s to low 300s based on testing.) ⭐ Deepseek At least 15-20 won from the start of the story, making it the ultimate value option. It also follows and expresses the narrative well according to the prompt. ⚠️ As the story progresses, the detailed instructions listed in the lorebook (Diary Notebook, !debug, etc.) may cause the prose style to become bland.