diff --git a/apps/application/workflow/nodes/text_to_video_node/text_to_video_node.py b/apps/application/workflow/nodes/text_to_video_node/text_to_video_node.py
index 5fe0a47d5eb..66b6cf9f6c6 100644
--- a/apps/application/workflow/nodes/text_to_video_node/text_to_video_node.py
+++ b/apps/application/workflow/nodes/text_to_video_node/text_to_video_node.py
@@ -93,8 +93,8 @@ def execute(self):
         question = self.workflow_manage.generate_prompt(prompt)
         self.write_context("question", question)
 
-        message_list = [*history_message, question]
-        self.write_context("message_list", message_list)
+        # message_list = [*history_message, question]
+        # self.write_context("message_list", [{"content": m.content, "role": m.type} for m in message_list],)
         self.write_context("dialogue_type", dialogue_type)
         self.write_context("negative_prompt", self.workflow_manage.generate_prompt(negative_prompt))
 
@@ -119,7 +119,7 @@ def execute(self):
 
         self.write_context("answer", video_label)
         self.write_context("video", video_list)
-        self.write_context("chat_model", ttv_model)
+        # self.write_context("chat_model", ttv_model)
 
         if is_result:
             node_info = NodeInfo(self.get_node_id(), self.get_node_name(), Status.SUCCESS)
diff --git a/ui/src/workflow-canvas/icons/text-to-video-node-icon.vue b/ui/src/workflow-canvas/icons/text-to-video-node-icon.vue
new file mode 100644
index 00000000000..7ea4a139f8c
--- /dev/null
+++ b/ui/src/workflow-canvas/icons/text-to-video-node-icon.vue
@@ -0,0 +1,6 @@
+<template>
+  <el-avatar class="ai-avatar-gradient" shape="square">
+    <img src="@/assets/workflow/icon_ai_chat.svg" style="width: 75%" alt="" />
+  </el-avatar>
+</template>
+<script setup lang="ts"></script>
diff --git a/ui/src/workflow-canvas/nodes/text-to-video-node/index.ts b/ui/src/workflow-canvas/nodes/text-to-video-node/index.ts
new file mode 100644
index 00000000000..c23206ba290
--- /dev/null
+++ b/ui/src/workflow-canvas/nodes/text-to-video-node/index.ts
@@ -0,0 +1,15 @@
+import TextToVideoNodeVue from './index.vue'
+import { WorkflowNodeModel, WorkflowNodeView } from '@/workflow-canvas/core/workflow-node'
+import { WorkflowNodeType } from '@/workflow-canvas/types.ts'
+
+class TextToVideoNodeView extends WorkflowNodeView {
+  constructor(props: ConstructorParameters<typeof WorkflowNodeView>[0]) {
+    super(props, TextToVideoNodeVue)
+  }
+}
+
+export default {
+  type: WorkflowNodeType.TextToVideoGenerateNode,
+  model: WorkflowNodeModel,
+  view: TextToVideoNodeView,
+}
diff --git a/ui/src/workflow-canvas/nodes/text-to-video-node/index.vue b/ui/src/workflow-canvas/nodes/text-to-video-node/index.vue
new file mode 100644
index 00000000000..52b20250d81
--- /dev/null
+++ b/ui/src/workflow-canvas/nodes/text-to-video-node/index.vue
@@ -0,0 +1,170 @@
+<script setup lang="ts">
+import { computed, inject, onMounted, ref, useTemplateRef } from 'vue'
+import { set } from 'lodash'
+import { QuestionFilled } from '@element-plus/icons-vue'
+import type { FormInstance } from 'element-plus'
+import ModelSelect from '@/components/business/model-select/index.vue'
+import NodeCascader from '@/workflow-canvas/core/NodeCascader.vue'
+import NodeContainer from '@/workflow-canvas/core/NodeContainer.vue'
+import { useWorkflowStore } from '@/workflow-canvas/store'
+import type { BaseNodeModel } from '@logicflow/core'
+import type { ModelItem, ModelProviderItem } from '@/api/types'
+
+defineOptions({ name: 'WorkflowTextToVideoNode' })
+const getModel = inject('getModel') as () => BaseNodeModel
+const apiType = (inject('apiType') as string) || 'workspace'
+const model = getModel()
+
+interface TextToVideoNodeForm {
+  model_id: string
+  model_id_type: 'custom' | 'reference'
+  model_id_reference: string[]
+  prompt: string
+  negative_prompt: string
+  model_params_setting: Record<string, unknown>
+}
+
+const formRef = useTemplateRef<FormInstance>('formRef')
+const modelCascaderRef = useTemplateRef<InstanceType<typeof NodeCascader>>('modelCascaderRef')
+
+const store = useWorkflowStore(apiType)
+const modelList = ref<Array<ModelItem>>([])
+const providerOptions = ref<Array<ModelProviderItem>>([])
+
+const formData = computed<TextToVideoNodeForm>({
+  get: () => {
+    if (!model.properties.node_data) {
+      set(model.properties, 'node_data', {
+        model_id: '',
+        model_id_type: 'custom',
+        model_id_reference: [],
+        prompt: '{{开始.question}}',
+        negative_prompt: '',
+        model_params_setting: {},
+      })
+    }
+    const data = model.properties.node_data as TextToVideoNodeForm
+    if (data.model_id_type === undefined) set(data, 'model_id_type', 'custom')
+    if (!Array.isArray(data.model_id_reference)) set(data, 'model_id_reference', [])
+    if (data.prompt === undefined) set(data, 'prompt', '{{开始.question}}')
+    if (data.negative_prompt === undefined) set(data, 'negative_prompt', '')
+    if (!data.model_params_setting) set(data, 'model_params_setting', {})
+    return data
+  },
+  set: (value) => (model.properties.node_data = value),
+})
+
+function validate() {
+  return Promise.all([
+    formData.value.model_id_type === 'reference'
+      ? modelCascaderRef.value?.validate()
+      : Promise.resolve(),
+    formRef.value?.validate(),
+  ]).catch((error) => Promise.reject({ node: model, errMessage: error }))
+}
+
+onMounted(() => {
+  set(model, 'validate', validate)
+  store.getModelList({ model_type: 'TTV' }).then((data) => {
+    modelList.value = data
+  })
+  store.getProviderList().then((data) => {
+    providerOptions.value = data
+  })
+})
+</script>
+
+<template>
+  <NodeContainer :node-model="model">
+    <h6 class="mb-3">节点设置</h6>
+    <el-form
+      ref="formRef"
+      :model="formData"
+      label-position="top"
+      require-asterisk-position="right"
+      @submit.prevent
+    >
+      <el-form-item
+        :prop="formData.model_id_type === 'reference' ? 'model_id_reference' : 'model_id'"
+        :rules="{ required: true, message: '请选择或填写文生视频模型', trigger: 'change' }"
+      >
+        <template #label>
+          <div class="flex-between gap-3 w-full">
+            <span>文生视频模型</span>
+            <el-select
+              v-model="formData.model_id_type"
+              :teleported="false"
+              class="w-30!"
+              size="small"
+              @change="formData.model_id_reference = []"
+            >
+              <el-option label="引用变量" value="reference" />
+              <el-option label="自定义" value="custom" />
+            </el-select>
+          </div>
+        </template>
+        <NodeCascader
+          v-if="formData.model_id_type === 'reference'"
+          ref="modelCascaderRef"
+          v-model="formData.model_id_reference"
+          :node-model="model"
+          class="w-full"
+          placeholder="请选择变量"
+        />
+        <ModelSelect
+          v-else
+          placeholder="请输入文生视频模型 ID"
+          :options="modelList"
+          :provider-options="providerOptions"
+          v-model="formData.model_id"
+        ></ModelSelect>
+      </el-form-item>
+
+      <el-form-item
+        label="提示词(正向)"
+        prop="prompt"
+        :rules="{ required: true, message: '请输入正向提示词', trigger: 'blur' }"
+      >
+        <template #label>
+          <div class="flex items-center gap-1">
+            <span>提示词(正向)</span>
+            <el-tooltip effect="dark" placement="right" content="正向提示词，用来描述期望生成的视频中应包含的元素和视觉特点。">
+              <MkIcon :icon="QuestionFilled" class="cursor-help text-N600" />
+            </el-tooltip>
+          </div>
+        </template>
+        <el-input
+          v-model="formData.prompt"
+          :rows="5"
+          :placeholder="`正向提示词，可以引用变量，如 {{开始.question}}`"
+          type="textarea"
+        />
+      </el-form-item>
+
+      <el-form-item label="提示词(负向)" prop="negative_prompt">
+        <template #label>
+          <div class="flex items-center gap-1">
+            <span>提示词(负向)</span>
+            <el-tooltip effect="dark" placement="right" content="反向提示词，用来描述不希望在画面中看到的内容，可以对画面进行限制。">
+              <MkIcon :icon="QuestionFilled" class="cursor-help text-N600" />
+            </el-tooltip>
+          </div>
+        </template>
+        <el-input
+          v-model="formData.negative_prompt"
+          :rows="4"
+          placeholder="请描述不想生成的图片内容，比如：颜色、血腥内容"
+          type="textarea"
+        />
+      </el-form-item>
+    </el-form>
+  </NodeContainer>
+</template>
+<style lang="scss" scoped>
+:deep(.el-form-item__label) {
+  width: 100%;
+}
+:deep(.el-form-item) {
+  margin-bottom: 16px;
+}
+</style>