> Handling Continuous Conversations Now, let’s improve our script to *remember past messages* and maintain a conversation. 📌 *Steps:* 1️⃣ Save *previous messages* in an array. 2️⃣ *Add user input* dynamically. 3️⃣ *Push AI responses* (excluding `reasoning_content`) back into the conversation. 📝 *Updated Code:* ``` const axios = require('axios'); const api_key = "your-api-key"; const url = "https://api.segmind.com/v1/deepseek-reasoner"; const messages = []; // Store conversation history async function sendMessage(userMessage) { // Add user message to array messages.push({ role: "user", content: userMessage }); try { const response = await axios.post(url, { messages }, { headers: { 'x-api-key': api_key } }); // Extract assistant response (ignore reasoning_content) const aiResponse = response.data.choices[0]?.message?.content || "No response received."; // Save AI response to conversation messages.push({ role: "assistant", content: aiResponse }); console.log("AI:", aiResponse); } catch (error) { console.log("Error:", error.response?.data || error.message); } } sendMessage("Hello, what can you do?"); setTimeout(() => sendMessage("Tell me a joke!"), 20000); ``` 🔄 *How It Works:* - The *messages array* keeps track of past interactions. - *New user messages* are added before sending the request. - *AI responses* are saved (excluding `reasoning_content`). - This allows a *smooth, context-aware chat experience*.