diff --git a/plugin/custom/plugins/CompleteWithLLM.js b/plugin/custom/plugins/CompleteWithLLM.js
new file mode 100644
index 00000000..812161e5
--- /dev/null
+++ b/plugin/custom/plugins/CompleteWithLLM.js
@@ -0,0 +1,95 @@
+// ./plugin/custom/plugins/CompleteWithLLM.js
+
+class CompleteWithLLM extends BaseCustomPlugin {
+    // 定义插件的快捷键
+    hotkey = () => [this.config.hotkey_string]
+
+    // 定义插件的提示信息
+    hint = () => "使用 LLM 补全选中的文本"
+
+    // 定义插件的初始化逻辑
+    init = () => {
+        // 初始化变量
+        this.selectedText = "";
+    }
+
+    // 定义插件的样式
+    style = () => `
+        #complete-with-llm-plugin {
+            margin: 10px;
+        }
+    `
+
+    // 定义插件的 HTML 结构
+    html = () => "<div id='complete-with-llm-plugin'></div>"
+
+    // 定义插件的处理逻辑
+    process = () => {
+        // 这里可以添加一些初始化逻辑
+    }
+
+    callback = async anchorNode =>  {
+        // 获取用户选中的文本
+        this.selectedText = window.getSelection().toString().replace(/\n\n+/g, '\n');
+
+        if (!this.selectedText) {
+            alert("请先选中一段文字");
+            return;
+        }
+
+        try {
+            const newContent = await this.sendToLLM(this.selectedText);
+
+            // 将新内容插入到原内容的下一行
+            console.log(`inserting the newContent ${newContent}`)
+            // this.insertNewContent(newContent);
+            this.utils.insertText(anchorNode, newContent)
+        } catch (error) {
+            console.error("LLM API 请求失败:", error);
+            alert("LLM API 请求失败，请检查网络或 API 配置");
+        }
+    }
+
+    sendToLLM = async function(text) {
+        const {apiKey, url, model, prompt} = this.config;
+
+        const data = {
+            model: model,
+            messages: [
+                {"role": "system", "content": prompt},
+                {"role": "user", "content": `input text: ${text}`}
+            ],
+            stream: false,
+            temperature: 0.01
+        };
+    
+        try {
+            console.log(JSON.stringify(data))
+            const response = await fetch(url, {
+                method: "POST",
+                headers: {
+                    "Content-Type": "application/json",
+                    "Authorization": `Bearer ${apiKey}`
+                },
+                body: JSON.stringify(data),
+            });
+    
+            if (!response.ok) {
+                throw new Error(`HTTP error! status: ${response.status}`);
+            }
+            const responseData = await response.json();
+            if (responseData.choices && responseData.choices.length > 0) {
+                return responseData.choices[0].message.content;
+            } else {
+                return "No response content found.";
+            }
+        } catch (error) {
+            console.error("Error in chatting with LLM:", error.message);
+            return `Error: ${error.message}`;
+        }
+    }
+
+}
+
+// 导出插件
+module.exports = { plugin: CompleteWithLLM };
\ No newline at end of file
diff --git a/plugin/global/settings/custom_plugin.user.toml b/plugin/global/settings/custom_plugin.user.toml
index e69de29b..745e8aa9 100644
--- a/plugin/global/settings/custom_plugin.user.toml
+++ b/plugin/global/settings/custom_plugin.user.toml
@@ -0,0 +1,15 @@
+[CompleteWithLLM]
+name = "使用LLM补全"
+enable = true
+hide = false
+order = 1
+hotkey_string = "ctrl+shift+H"
+apiKey = "sk-xxx"
+url = "https://api.deepseek.com"
+model = "deepseek-chat"
+prompt = """
+Please read the input text, and follow these instructions:\r
+1. Optimize and complete the input text. \r
+2. The language of the output should be consistent with the language of the input text, even if it is in a mixed-language form.\r
+3. Don't provide any additional explanation before or after the answer and output directly.\r
+"""
\ No newline at end of file