diff --git a/components/Conversation.vue b/components/Conversation.vue
index 99140e4..5254afb 100644
--- a/components/Conversation.vue
+++ b/components/Conversation.vue
@@ -180,6 +180,10 @@ watchEffect(() => {
   }
 })
 
+onNuxtReady(() => {
+  currentModel.value = getCurrentModel()
+})
+
 </script>
 
 <template>
diff --git a/components/ModelParameters.vue b/components/ModelParameters.client.vue
similarity index 88%
rename from components/ModelParameters.vue
rename to components/ModelParameters.client.vue
index 69c1e88..6a119e4 100644
--- a/components/ModelParameters.vue
+++ b/components/ModelParameters.client.vue
@@ -8,10 +8,13 @@ const availableModels = [
 ]
 const currentModelDefault = ref(MODELS[currentModel.value.name])
 
-watch(currentModel, (newVal, oldVal) => {
-  currentModelDefault.value = MODELS[newVal.name]
-  saveCurrentModel(newVal)
-}, { deep: true })
+onNuxtReady(() => {
+  currentModel.value = getCurrentModel()
+  watch(currentModel, (newVal, oldVal) => {
+    currentModelDefault.value = MODELS[newVal.name]
+    saveCurrentModel(newVal)
+  }, { deep: true })
+})
 
 </script>
 
@@ -53,7 +56,7 @@ watch(currentModel, (newVal, oldVal) => {
             <div class="d-flex justify-space-between align-center">
               <v-list-subheader>{{ $t('temperature') }}</v-list-subheader>
               <v-text-field
-                  v-model="currentModel.temperature"
+                  v-model.number="currentModel.temperature"
                   hide-details
                   single-line
                   density="compact"
@@ -82,7 +85,7 @@ watch(currentModel, (newVal, oldVal) => {
             <div class="d-flex justify-space-between align-center">
               <v-list-subheader>{{ $t('maxTokens') }}</v-list-subheader>
               <v-text-field
-                  v-model="currentModel.max_tokens"
+                  v-model.number="currentModel.max_tokens"
                   hide-details
                   single-line
                   density="compact"
@@ -93,6 +96,9 @@ watch(currentModel, (newVal, oldVal) => {
                   class="flex-grow-0"
               ></v-text-field>
             </div>
+            <div class="text-caption">
+              {{ $t('maxTokenTips1') }} <b>{{ currentModelDefault.total_tokens }}</b> {{ $t('maxTokenTips2') }}
+            </div>
           </v-col>
           <v-col cols="12">
             <v-slider
@@ -111,7 +117,7 @@ watch(currentModel, (newVal, oldVal) => {
             <div class="d-flex justify-space-between align-center">
               <v-list-subheader>{{ $t('topP') }}</v-list-subheader>
               <v-text-field
-                  v-model="currentModel.top_p"
+                  v-model.number="currentModel.top_p"
                   hide-details
                   single-line
                   density="compact"
@@ -138,7 +144,7 @@ watch(currentModel, (newVal, oldVal) => {
             <div class="d-flex justify-space-between align-center">
               <v-list-subheader>{{ $t('frequencyPenalty') }}</v-list-subheader>
               <v-text-field
-                  v-model="currentModel.frequency_penalty"
+                  v-model.number="currentModel.frequency_penalty"
                   hide-details
                   single-line
                   density="compact"
@@ -164,7 +170,7 @@ watch(currentModel, (newVal, oldVal) => {
             <div class="d-flex justify-space-between align-center">
               <v-list-subheader>{{ $t('presencePenalty') }}</v-list-subheader>
               <v-text-field
-                  v-model="currentModel.presence_penalty"
+                  v-model.number="currentModel.presence_penalty"
                   hide-details
                   single-line
                   density="compact"
diff --git a/lang/en-US.json b/lang/en-US.json
index 3908fb3..a3921a1 100644
--- a/lang/en-US.json
+++ b/lang/en-US.json
@@ -50,6 +50,8 @@
   "webSearch": "Web Search",
   "webSearchDefaultPrompt": "Web search results:\n\n[web_results]\nCurrent date: [current_date]\n\nInstructions: Using the provided web search results, write a comprehensive reply to the given query. Make sure to cite results using [[number](URL)] notation after the reference. If the provided search results refer to multiple subjects with the same name, write separate answers for each subject.\nQuery: [query]",
   "genTitlePrompt": "Generate a short title for the following content, no more than 10 words. \n\nContent: ",
+  "maxTokenTips1": "The maximum context length of the current model is",
+  "maxTokenTips2": "token, which includes the length of the prompt and the length of the generated text. The `Max Tokens` here refers to the length of the generated text. Therefore, you should leave some space for your prompt and not set it too large or to the maximum.",
   "welcomeScreen": {
     "introduction1": "is an unofficial client for ChatGPT, but uses the official OpenAI API.",
     "introduction2": "You will need an OpenAI API Key before you can use this client.",
diff --git a/lang/ru-RU.json b/lang/ru-RU.json
index 7f1bd48..a27ad20 100644
--- a/lang/ru-RU.json
+++ b/lang/ru-RU.json
@@ -50,6 +50,8 @@
   "webSearch": "Поиск в интернете",
   "webSearchDefaultPrompt": "Результаты веб-поиска:\n\n[web_results]\nТекущая дата: [current_date]\n\nИнструкции: Используя предоставленные результаты веб-поиска, напишите развернутый ответ на заданный запрос. Обязательно цитируйте результаты, используя обозначение [[number](URL)] после ссылки. Если предоставленные результаты поиска относятся к нескольким темам с одинаковым названием, напишите отдельные ответы для каждой темы.\nЗапрос: [query]",
   "genTitlePrompt": "Придумайте короткий заголовок для следующего содержания, не более 10 слов. \n\nСодержание: ",
+  "maxTokenTips1": "The maximum context length of the current model is",
+  "maxTokenTips2": "token, which includes the length of the prompt and the length of the generated text. The `Max Tokens` here refers to the length of the generated text. Therefore, you should leave some space for your prompt and not set it too large or to the maximum.",
   "welcomeScreen": {
     "introduction1": "является неофициальным клиентом для ChatGPT, но использует официальный API OpenAI.",
     "introduction2": "Вам понадобится ключ API OpenAI, прежде чем вы сможете использовать этот клиент.",
diff --git a/lang/zh-CN.json b/lang/zh-CN.json
index b2ff081..67e2dd7 100644
--- a/lang/zh-CN.json
+++ b/lang/zh-CN.json
@@ -50,6 +50,8 @@
   "webSearch": "网页搜索",
   "webSearchDefaultPrompt": "网络搜索结果：\n\n[web_results]\n当前日期：[current_date]\n\n说明：使用提供的网络搜索结果，对给定的查询写出全面的回复。确保在引用参考文献后使用 [[number](URL)] 符号进行引用结果. 如果提供的搜索结果涉及到多个具有相同名称的主题，请针对每个主题编写单独的答案。\n查询：[query]",
   "genTitlePrompt": "为以下内容生成一个不超过10个字的简短标题。 \n\n内容: ",
+  "maxTokenTips1": "当前模型的最大上下文长度为",
+  "maxTokenTips2": "个 token，它包括了指令的长度和生成的文本长度。此处的最大 token 数量是指生成的文本长度。所以您应该为您的指令预留一些空间，不宜设置过大或拉满。",
   "welcomeScreen": {
     "introduction1": "是一个非官方的ChatGPT客户端，但使用OpenAI的官方API",
     "introduction2": "在使用本客户端之前，您需要一个OpenAI API密钥。",