feat: add target model to messages so we dont have to peek into child ai message responses every time to know what model the user sent their message to for tokenValue computations

api/server/services/Threads/manage.js

@@ -88,6 +88,7 @@ async function saveUserMessage(req, params) {
     parentMessageId: params.parentMessageId ?? Constants.NO_PARENT,
     /* For messages, use the assistant_id instead of model */
     model: params.assistant_id,
+    targetModel: params.model,
     thread_id: params.thread_id,
     sender: 'User',
     text: params.text,