diff --git a/.aider.conf.yml b/.aider.conf.yml
index 8619221..aa77361 100644
--- a/.aider.conf.yml
+++ b/.aider.conf.yml
@@ -1 +1 @@
-model: ollama_chat/qwen2.5-coder:14b-instruct-q4_K_M
\ No newline at end of file
+model: /models/Qwen/Qwen2.5-Coder-32B-Instruct-GGUF/qwen2.5-coder-32b-instruct-q4_k_m.gguf
\ No newline at end of file
diff --git a/.aider.model.settings.yml b/.aider.model.settings.yml
index 43facbf..0e3af4d 100644
--- a/.aider.model.settings.yml
+++ b/.aider.model.settings.yml
@@ -1,3 +1,4 @@
-- name: /models/qwen2.5-coder-32b-instruct-q4_k_m.gguf
+- name: /models/Qwen/Qwen2.5-Coder-32B-Instruct-GGUF/qwen2.5-coder-32b-instruct-q4_k_m.gguf
   extra_params:
-    num_ctx: 16384
\ No newline at end of file
+    num_ctx: 16384
+    num_threads: 8
\ No newline at end of file
diff --git a/.aiignore b/.aiignore
index 62c8935..a3a7798 100644
--- a/.aiignore
+++ b/.aiignore
@@ -1 +1,5 @@
-.idea/
\ No newline at end of file
+.idea/
+.aider.tags.cache.v4
+.aider.chat.history.md
+.aider.input.history
+.env
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..40816a2
--- /dev/null
+++ b/README.md
@@ -0,0 +1 @@
+Hi
\ No newline at end of file
diff --git a/chat.js b/chatv2.js
similarity index 85%
rename from chat.js
rename to chatv2.js
index 76e5927..be84359 100644
--- a/chat.js
+++ b/chatv2.js
@@ -2,23 +2,24 @@
 const BACKENDS = {
 	plato : {
 		prod: '/api/plato',
-		dev : 'http://192.168.1.74:1234/v1',
+		dev : 'http://192.168.1.74:1234/v1', // Removed trailing space
 		name: 'Plato (192.168.1.74)'
 	},
 	stoic : {
 		prod: '/api/stoic',
-		dev : 'http://192.168.1.159:1234/v1',
+		dev : 'http://192.168.1.159:1234/v1', // Removed trailing space
 		name: 'Stoic (192.168.1.159)'
 	},
 	ollama: {
 		prod: '/api/ollama',
-		dev : 'http://192.168.1.159:8081/v1',
+		dev : 'http://192.168.1.159:8081/v1', // Removed trailing space
 		name: 'Ollama (192.168.1.159:8081)'
 	}
 }
 
 const IS_PRODUCTION = window.location.hostname === 'jarvis-lan.appmodel.nl'
 const API_KEY = 'not-needed'
+const MAX_CHAT_HISTORY = 50 // Maximum messages to keep in history
 
 // Global state
 let currentBackend = 'plato'
@@ -26,7 +27,7 @@ let currentModel = null
 let availableModels = []
 let currentStreamController = null
 let isStreaming = false
-let chatHistory = [] // Initialize chat history array
+let chatHistory = [] // Stores conversation history
 
 // Get current API URL based on selected backend
 function getApiUrl() {
@@ -38,7 +39,7 @@ function getApiUrl() {
 function updateBackendDisplay() {
 	const backend = BACKENDS[currentBackend]
 	const displayText = IS_PRODUCTION
-		? `${ backend.prod } → ${ backend.name }`
+		? `${ backend.prod } ? ${ backend.name }`
 		: backend.dev
 	document.getElementById('backendDisplay').textContent = displayText
 }
@@ -84,6 +85,23 @@ function populateModelSelector() {
 	}
 }
 
+// Get formatted messages for API (excludes UI-specific properties)
+function getApiMessages() {
+	// Get last N messages and format for API
+	return chatHistory.slice(-MAX_CHAT_HISTORY).map(({ role, content }) => ({
+		role,
+		content
+	}))
+}
+
+// Trim chat history if it exceeds maximum
+function trimChatHistory() {
+	if (chatHistory.length > MAX_CHAT_HISTORY) {
+		chatHistory = chatHistory.slice(-MAX_CHAT_HISTORY)
+		console.log(`Chat history trimmed to ${ MAX_CHAT_HISTORY } messages`)
+	}
+}
+
 // Handle backend and model selection changes
 document.addEventListener('DOMContentLoaded', () => {
 	const backendSelector = document.getElementById('backendSelector')
@@ -96,7 +114,7 @@ document.addEventListener('DOMContentLoaded', () => {
 	backendSelector.addEventListener('change', (e) => {
 		currentBackend = e.target.value
 		updateBackendDisplay()
-		console.log('Backend switched to:', currentBackend, '→', getApiUrl())
+		console.log('Backend switched to:', currentBackend, '?', getApiUrl())
 		fetchModels() // Reload models for new backend
 	})
 
@@ -146,7 +164,7 @@ function addMessage(role, content, markdown = false, messageId = null) {
 
 	const avatar = document.createElement('div')
 	avatar.className = `avatar ${ role }-avatar`
-	avatar.textContent = role === 'user' ? '👤' : '🤖'
+	avatar.textContent = role === 'user' ? '?' : '?'
 
 	const name = document.createElement('span')
 	name.textContent = role === 'user' ? 'You' : 'Assistant'
@@ -183,9 +201,12 @@ function addMessage(role, content, markdown = false, messageId = null) {
 	chatLog.appendChild(messageDiv)
 	chatLog.scrollTop = chatLog.scrollHeight
 
-	// Add message to chat history
+	// Store message in history (excluding UI-specific properties)
 	chatHistory.push({ role, content, markdown, messageId })
 
+	// Trim history if it gets too long
+	trimChatHistory()
+
 	return contentDiv
 }
 
@@ -201,8 +222,8 @@ function parseThinkingTags(content) {
 		const id = `thinking-${ Date.now() }-${ thinkingCounter }`
 		return `<div class="thinking-block">
 			<div class="thinking-header" onclick="toggleThinking('${ id }')">
-				<span class="thinking-toggle" id="${ id }-toggle">▼</span>
-				<span>🤔 Thinking...</span>
+				<span class="thinking-toggle" id="${ id }-toggle">?</span>
+				<span>? Thinking...</span>
 			</div>
 			<div class="thinking-content" id="${ id }">${ thinkContent.trim() }</div>
 		</div>`
@@ -215,8 +236,8 @@ function parseThinkingTags(content) {
 		const id = `thinking-${ Date.now() }-${ thinkingCounter }`
 		return `<div class="thinking-block">
 			<div class="thinking-header" onclick="toggleThinking('${ id }')">
-				<span class="thinking-toggle" id="${ id }-toggle">▼</span>
-				<span>🤔 Thinking...</span>
+				<span class="thinking-toggle" id="${ id }-toggle">?</span>
+				<span>? Thinking...</span>
 			</div>
 			<div class="thinking-content" id="${ id }">${ thinkContent.trim() }</div>
 		</div>`
@@ -267,7 +288,7 @@ function showTypingIndicator() {
 
 	const avatar = document.createElement('div')
 	avatar.className = 'avatar assistant-avatar'
-	avatar.textContent = '🤖'
+	avatar.textContent = '?'
 
 	const name = document.createElement('span')
 	name.textContent = 'Assistant'
@@ -315,7 +336,7 @@ async function handleStreamingResponse(userMessage) {
 
 	const avatar = document.createElement('div')
 	avatar.className = 'avatar assistant-avatar'
-	avatar.textContent = '🤖'
+	avatar.textContent = '?'
 
 	const name = document.createElement('span')
 	name.textContent = 'Assistant'
@@ -344,7 +365,7 @@ async function handleStreamingResponse(userMessage) {
 
 		const requestBody = {
 			model      : currentModel || 'local-model',
-			messages   : [{ role: 'user', content: userMessage }],
+			messages   : getApiMessages(), // Send full chat history
 			stream     : true,
 			temperature: 0.7,
 			max_tokens : 2000
@@ -384,7 +405,7 @@ async function handleStreamingResponse(userMessage) {
 				// Update immediately if enough time has passed
 				updateScheduled = true
 				requestAnimationFrame(() => {
-					updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
+					updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked, 'assistant')
 					if (cursorSpan.parentNode) {
 						contentDiv.appendChild(cursorSpan)
 					}
@@ -397,7 +418,7 @@ async function handleStreamingResponse(userMessage) {
 				updateScheduled = true
 				setTimeout(() => {
 					requestAnimationFrame(() => {
-						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
+						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked, 'assistant')
 						if (cursorSpan.parentNode) {
 							contentDiv.appendChild(cursorSpan)
 						}
@@ -424,7 +445,7 @@ async function handleStreamingResponse(userMessage) {
 						isStreaming = false
 						currentStreamController = null
 						// Final update and remove cursor
-						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
+						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked, 'assistant')
 						if (cursorSpan.parentNode) {
 							cursorSpan.parentNode.removeChild(cursorSpan)
 						}
@@ -446,19 +467,18 @@ async function handleStreamingResponse(userMessage) {
 		}
 
 		// Final update
-		updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
+		updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked, 'assistant')
 		chatLog.scrollTop = chatLog.scrollHeight
 
 	} catch (error) {
 		console.error('Streaming error:', error)
 		if (error.name !== 'AbortError') {
-			updateMessageContent(contentDiv, `Error: ${ error.message }`, false)
+			updateMessageContent(contentDiv, `Error: ${ error.message }`, false, 'assistant')
 		}
 	} finally {
 		isStreaming = false
 		currentStreamController = null
 		sendBtn.disabled = false
-		userInput.focus()
 
 		// Remove cursor if still exists
 		if (cursorSpan.parentNode) {
@@ -480,7 +500,7 @@ async function handleNonStreamingResponse(userMessage) {
 
 		const requestBody = {
 			model      : currentModel || 'local-model',
-			messages   : [{ role: 'user', content: userMessage }],
+			messages   : getApiMessages(), // Send full chat history
 			stream     : false,
 			temperature: 0.7,
 			max_tokens : 2000
@@ -519,12 +539,6 @@ async function handleNonStreamingResponse(userMessage) {
 	}
 }
 
-function loadChatHistory() {
-	chatHistory.forEach(message => {
-		addMessage(message.role, message.content, message.markdown, message.messageId)
-	})
-}
-
 // Stop current stream
 function stopStream() {
 	if (currentStreamController) {
@@ -541,17 +555,24 @@ async function sendMessage() {
 	const userMessage = userInput.value.trim()
 	if (!userMessage || isStreaming) return
 
+	// Add user message to UI and history
 	addMessage('user', userMessage)
+
+	// Clear input field
 	userInput.value = ''
 	userInput.style.height = 'auto'
 	sendBtn.disabled = true
 
+	// Update button text based on streaming state
 	if (streamToggle.checked) {
 		sendBtn.textContent = 'Stop'
-		handleStreamingResponse(userMessage)
+		await handleStreamingResponse(userMessage)
 	} else {
-		handleNonStreamingResponse(userMessage)
+		await handleNonStreamingResponse(userMessage)
 	}
+
+	// Reset button text
+	sendBtn.textContent = 'Send'
 }
 
 // Event Listeners
@@ -577,28 +598,25 @@ window.onload = () => {
 	userInput.focus()
 	fetchModels()
 
-	// Load chat history
-	loadChatHistory()
+	// Add welcome message (this will be the first entry in chatHistory)
+	const welcomeMessage = `# Welcome to LM Studio Chat! ?
 
-	// Add welcome message
-	setTimeout(() => {
-		const welcomeMessage = `# Welcome to LM Studio Chat! 🚀
-
-I now support **real-time streaming responses** and **dark, readable text formatting**.
+I now support **full conversation history**, **real-time streaming responses** and **dark, readable text formatting**.
 
 ## Features:
-1. **Streaming Mode** (enabled by default) - Watch responses appear word-by-word
-2. **Markdown Rendering** - Proper formatting for code, lists, tables, and more
-3. **Readable Dark Text** - No more eye strain from light gray text
-4. **Model Detection** - Automatically shows your loaded model
+1. **Full Chat History** - Complete conversation context is now sent to the AI
+2. **Smart History Management** - Automatically keeps the last ${ MAX_CHAT_HISTORY } messages
+3. **Streaming Mode** (enabled by default) - Watch responses appear word-by-word
+4. **Markdown Rendering** - Proper formatting for code, lists, tables, and more
+5. **Readable Dark Text** - No more eye strain from light gray text
 
 ## Try it out:
-- Ask a technical question to see **code formatting**
-- Type \`ls -la\` in a code block
-- Request a **numbered list** or table
+- Ask follow-up questions that reference earlier messages
+- Have a multi-turn conversation with full context
+- Ask "what did I just ask?" to test history retention
 - Watch the streaming response in real-time!
 
 > *Tip: You can toggle streaming and markdown using the checkboxes below.*`
-		addMessage('assistant', welcomeMessage, true)
-	}, 500)
-}
+
+	addMessage('assistant', welcomeMessage, true)
+}
\ No newline at end of file
diff --git a/index.html b/index.html
index 72850e2..9034621 100644
--- a/index.html
+++ b/index.html
@@ -58,612 +58,6 @@
     </div>
 </div>
 
-<script>
-// Configuration
-const BACKENDS = {
-	plato : {
-		prod: '/api/plato',
-		dev : 'http://192.168.1.74:1234/v1',
-		name: 'Plato (192.168.1.74)'
-	},
-	stoic : {
-		prod: '/api/stoic',
-		dev : 'http://192.168.1.159:1234/v1',
-		name: 'Stoic (192.168.1.159)'
-	},
-	ollama: {
-		prod: '/api/ollama',
-		dev : 'http://192.168.1.159:8081/v1',
-		name: 'Ollama (192.168.1.159:8081)'
-	}
-}
-
-const IS_PRODUCTION = window.location.hostname === 'jarvis-lan.appmodel.nl'
-const API_KEY = 'not-needed'
-
-// Global state
-let currentBackend = 'plato'
-let currentModel = null
-let availableModels = []
-let currentStreamController = null
-let isStreaming = false
-let chatHistory = [] // Initialize chat history array
-
-// Get current API URL based on selected backend
-function getApiUrl() {
-	const backend = BACKENDS[currentBackend]
-	return IS_PRODUCTION ? backend.prod : backend.dev
-}
-
-// Update backend display
-function updateBackendDisplay() {
-	const backend = BACKENDS[currentBackend]
-	const displayText = IS_PRODUCTION
-		? `${ backend.prod } → ${ backend.name }`
-		: backend.dev
-	document.getElementById('backendDisplay').textContent = displayText
-}
-
-// Fetch available models from backend
-async function fetchModels() {
-	try {
-		const response = await fetch(`${ getApiUrl() }/models`)
-		if (response.ok) {
-			const data = await response.json()
-			availableModels = data.data || []
-			populateModelSelector()
-
-			// Auto-select first model if none selected
-			if (!currentModel && availableModels.length > 0) {
-				currentModel = availableModels[0].id
-				document.getElementById('modelSelector').value = currentModel
-			}
-		} else {
-			console.error('Failed to fetch models:', response.statusText)
-			document.getElementById('modelSelector').innerHTML = '<option value="">Error loading models</option>'
-		}
-	} catch (error) {
-		console.error('Error fetching models:', error)
-		document.getElementById('modelSelector').innerHTML = '<option value="">Error loading models</option>'
-	}
-}
-
-// Populate model selector dropdown
-function populateModelSelector() {
-	const selector = document.getElementById('modelSelector')
-	if (availableModels.length === 0) {
-		selector.innerHTML = '<option value="">No models available</option>'
-		return
-	}
-
-	selector.innerHTML = availableModels.map(model =>
-		`<option value="${ model.id }">${ model.id }</option>`
-	).join('')
-
-	if (currentModel) {
-		selector.value = currentModel
-	}
-}
-
-// Handle backend and model selection changes
-document.addEventListener('DOMContentLoaded', () => {
-	const backendSelector = document.getElementById('backendSelector')
-	const modelSelector = document.getElementById('modelSelector')
-
-	backendSelector.value = currentBackend
-	updateBackendDisplay()
-	fetchModels()
-
-	backendSelector.addEventListener('change', (e) => {
-		currentBackend = e.target.value
-		updateBackendDisplay()
-		console.log('Backend switched to:', currentBackend, '→', getApiUrl())
-		fetchModels() // Reload models for new backend
-	})
-
-	modelSelector.addEventListener('change', (e) => {
-		currentModel = e.target.value
-		console.log('Model selected:', currentModel)
-	})
-})
-
-// DOM Elements
-const chatLog = document.getElementById('chatLog')
-const userInput = document.getElementById('userInput')
-const sendBtn = document.getElementById('sendBtn')
-const streamToggle = document.getElementById('streamToggle')
-const markdownToggle = document.getElementById('markdownToggle')
-const modelNameSpan = document.getElementById('modelName')
-
-// Initialize marked with options
-marked.setOptions({
-	gfm      : true,
-	breaks   : true,
-	highlight: function(code, lang) {
-		if (window.hljs) {
-			const language = hljs.getLanguage(lang) ? lang : 'plaintext'
-			return hljs.highlight(code, { language }).value
-		}
-		return code
-	}
-})
-
-// Auto-resize textarea
-function autoResize(textarea) {
-	textarea.style.height = 'auto'
-	textarea.style.height = Math.min(textarea.scrollHeight, 300) + 'px'
-}
-
-// Old fetchModels function removed - now handled in DOMContentLoaded
-
-// Add message to chat
-function addMessage(role, content, markdown = false, messageId = null) {
-	const messageDiv = document.createElement('div')
-	messageDiv.className = `message message-${ role }`
-	if (messageId) {
-		messageDiv.id = messageId
-	}
-
-	const headerDiv = document.createElement('div')
-	headerDiv.className = `message-header ${ role }`
-
-	const avatar = document.createElement('div')
-	avatar.className = `avatar ${ role }-avatar`
-	avatar.textContent = role === 'user' ? '👤' : '🤖'
-
-	const name = document.createElement('span')
-	name.textContent = role === 'user' ? 'You' : 'Assistant'
-
-	headerDiv.appendChild(avatar)
-	headerDiv.appendChild(name)
-
-	const contentDiv = document.createElement('div')
-
-	if (markdown && role === 'assistant' && markdownToggle.checked) {
-		contentDiv.className = 'markdown-content'
-		// Parse thinking tags before markdown
-		const processedContent = parseThinkingTags(content)
-		contentDiv.innerHTML = marked.parse(processedContent)
-		// Apply syntax highlighting if hljs is available
-		if (window.hljs) {
-			setTimeout(() => {
-				contentDiv.querySelectorAll('pre code').forEach((block) => {
-					hljs.highlightElement(block)
-				})
-			}, 0)
-		}
-	} else {
-		// Parse thinking tags for non-markdown too
-		const processedContent = parseThinkingTags(content)
-		contentDiv.innerHTML = processedContent
-		contentDiv.style.whiteSpace = 'pre-wrap'
-		contentDiv.style.padding = '8px 0'
-		contentDiv.style.color = '#2d3339'
-	}
-
-	messageDiv.appendChild(headerDiv)
-	messageDiv.appendChild(contentDiv)
-	chatLog.appendChild(messageDiv)
-	chatLog.scrollTop = chatLog.scrollHeight
-
-	// Add message to chat history
-	chatHistory.push({ role, content, markdown, messageId })
-
-	return contentDiv
-}
-
-// Parse [THINK] tags and create collapsible sections
-function parseThinkingTags(content) {
-	let result = content
-	let thinkingCounter = 0
-
-	// Handle complete [THINK]...[/THINK] blocks
-	const completeThinkRegex = /\[THINK\]([\s\S]*?)\[\/THINK\]/gi
-	result = result.replace(completeThinkRegex, (match, thinkContent) => {
-		thinkingCounter++
-		const id = `thinking-${ Date.now() }-${ thinkingCounter }`
-		return `<div class="thinking-block">
-			<div class="thinking-header" onclick="toggleThinking('${ id }')">
-				<span class="thinking-toggle" id="${ id }-toggle">▼</span>
-				<span>🤔 Thinking...</span>
-			</div>
-			<div class="thinking-content" id="${ id }">${ thinkContent.trim() }</div>
-		</div>`
-	})
-
-	// Handle incomplete [THINK] blocks (still streaming)
-	const incompleteThinkRegex = /\[THINK\]([\s\S]*?)$/gi
-	result = result.replace(incompleteThinkRegex, (match, thinkContent) => {
-		thinkingCounter++
-		const id = `thinking-${ Date.now() }-${ thinkingCounter }`
-		return `<div class="thinking-block">
-			<div class="thinking-header" onclick="toggleThinking('${ id }')">
-				<span class="thinking-toggle" id="${ id }-toggle">▼</span>
-				<span>🤔 Thinking...</span>
-			</div>
-			<div class="thinking-content" id="${ id }">${ thinkContent.trim() }</div>
-		</div>`
-	})
-
-	return result
-}
-
-// Toggle thinking section visibility
-window.toggleThinking = function(id) {
-	const content = document.getElementById(id)
-	const toggle = document.getElementById(id + '-toggle')
-
-	if (content && toggle) {
-		content.classList.toggle('hidden')
-		toggle.classList.toggle('collapsed')
-	}
-}
-
-// Update message content (for streaming)
-function updateMessageContent(contentDiv, newContent, markdown = false) {
-	// Parse thinking tags first
-	const processedContent = parseThinkingTags(newContent)
-
-	if (markdown && markdownToggle.checked) {
-		contentDiv.innerHTML = marked.parse(processedContent)
-		if (window.hljs) {
-			setTimeout(() => {
-				contentDiv.querySelectorAll('pre code').forEach((block) => {
-					hljs.highlightElement(block)
-				})
-			}, 0)
-		}
-	} else {
-		contentDiv.textContent = newContent
-	}
-}
-
-// Show typing indicator
-function showTypingIndicator() {
-	const typingDiv = document.createElement('div')
-	typingDiv.id = 'typingIndicator'
-	typingDiv.className = 'message message-assistant'
-
-	const headerDiv = document.createElement('div')
-	headerDiv.className = 'message-header assistant'
-
-	const avatar = document.createElement('div')
-	avatar.className = 'avatar assistant-avatar'
-	avatar.textContent = '🤖'
-
-	const name = document.createElement('span')
-	name.textContent = 'Assistant'
-
-	headerDiv.appendChild(avatar)
-	headerDiv.appendChild(name)
-
-	const dotsDiv = document.createElement('div')
-	dotsDiv.className = 'typing-indicator'
-	dotsDiv.innerHTML = `
-                <div class="typing-dot"></div>
-                <div class="typing-dot"></div>
-                <div class="typing-dot"></div>
-                <span style="margin-left: 8px;">Thinking...</span>
-            `
-
-	typingDiv.appendChild(headerDiv)
-	typingDiv.appendChild(dotsDiv)
-	chatLog.appendChild(typingDiv)
-	chatLog.scrollTop = chatLog.scrollHeight
-
-	return typingDiv
-}
-
-// Remove typing indicator
-function removeTypingIndicator(typingDiv) {
-	if (typingDiv && typingDiv.parentNode) {
-		typingDiv.parentNode.removeChild(typingDiv)
-	}
-}
-
-// Handle streaming response
-async function handleStreamingResponse(userMessage) {
-	const messageId = 'msg-' + Date.now()
-	let accumulatedContent = ''
-	let contentDiv = null
-
-	// Create initial empty message
-	const messageDiv = document.createElement('div')
-	messageDiv.className = 'message message-assistant'
-	messageDiv.id = messageId
-
-	const headerDiv = document.createElement('div')
-	headerDiv.className = 'message-header assistant'
-
-	const avatar = document.createElement('div')
-	avatar.className = 'avatar assistant-avatar'
-	avatar.textContent = '🤖'
-
-	const name = document.createElement('span')
-	name.textContent = 'Assistant'
-
-	headerDiv.appendChild(avatar)
-	headerDiv.appendChild(name)
-
-	contentDiv = document.createElement('div')
-	contentDiv.className = 'markdown-content'
-
-	messageDiv.appendChild(headerDiv)
-	messageDiv.appendChild(contentDiv)
-	chatLog.appendChild(messageDiv)
-
-	// Add streaming cursor
-	const cursorSpan = document.createElement('span')
-	cursorSpan.className = 'streaming-cursor'
-	contentDiv.appendChild(cursorSpan)
-
-	try {
-		// Auto-select first model if none selected
-		if (!currentModel && availableModels.length > 0) {
-			currentModel = availableModels[0].id
-			document.getElementById('modelSelector').value = currentModel
-		}
-
-		const requestBody = {
-			model      : currentModel || 'local-model',
-			messages   : [{ role: 'user', content: userMessage }],
-			stream     : true,
-			temperature: 0.7,
-			max_tokens : 2000
-		}
-
-		const response = await fetch(`${ getApiUrl() }/chat/completions`, {
-			method : 'POST',
-			headers: {
-				'Content-Type' : 'application/json',
-				'Authorization': `Bearer ${ API_KEY }`
-			},
-			body   : JSON.stringify(requestBody)
-		})
-
-		if (!response.ok) {
-			throw new Error(`API Error: ${ response.status } ${ response.statusText }`)
-		}
-
-		const reader = response.body.getReader()
-		const decoder = new TextDecoder()
-
-		isStreaming = true
-		currentStreamController = new AbortController()
-
-		// Debounce DOM updates to reduce stutter
-		let updateScheduled = false
-		let lastUpdate = Date.now()
-		const MIN_UPDATE_INTERVAL = 100 // ms, update at most every 100ms
-
-		function scheduleUpdate() {
-			if (updateScheduled) return
-
-			const now = Date.now()
-			const timeSinceLastUpdate = now - lastUpdate
-
-			if (timeSinceLastUpdate >= MIN_UPDATE_INTERVAL) {
-				// Update immediately if enough time has passed
-				updateScheduled = true
-				requestAnimationFrame(() => {
-					updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
-					if (cursorSpan.parentNode) {
-						contentDiv.appendChild(cursorSpan)
-					}
-					chatLog.scrollTop = chatLog.scrollHeight
-					lastUpdate = Date.now()
-					updateScheduled = false
-				})
-			} else {
-				// Schedule update after remaining time
-				updateScheduled = true
-				setTimeout(() => {
-					requestAnimationFrame(() => {
-						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
-						if (cursorSpan.parentNode) {
-							contentDiv.appendChild(cursorSpan)
-						}
-						chatLog.scrollTop = chatLog.scrollHeight
-						lastUpdate = Date.now()
-						updateScheduled = false
-					})
-				}, MIN_UPDATE_INTERVAL - timeSinceLastUpdate)
-			}
-		}
-
-		while (true) {
-			const { done, value } = await reader.read()
-			if (done) break
-
-			const chunk = decoder.decode(value)
-			const lines = chunk.split('\n').filter(line => line.trim() !== '')
-
-			for (const line of lines) {
-				if (line.startsWith('data: ')) {
-					const data = line.slice(6)
-
-					if (data === '[DONE]') {
-						isStreaming = false
-						currentStreamController = null
-						// Final update and remove cursor
-						updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
-						if (cursorSpan.parentNode) {
-							cursorSpan.parentNode.removeChild(cursorSpan)
-						}
-						chatLog.scrollTop = chatLog.scrollHeight
-						return
-					}
-
-					try {
-						const parsed = JSON.parse(data)
-						if (parsed.choices && parsed.choices[0].delta.content) {
-							accumulatedContent += parsed.choices[0].delta.content
-							scheduleUpdate()
-						}
-					} catch (e) {
-						console.log('Error parsing stream data:', e)
-					}
-				}
-			}
-		}
-
-		// Final update
-		updateMessageContent(contentDiv, accumulatedContent, markdownToggle.checked)
-		chatLog.scrollTop = chatLog.scrollHeight
-
-	} catch (error) {
-		console.error('Streaming error:', error)
-		if (error.name !== 'AbortError') {
-			updateMessageContent(contentDiv, `Error: ${ error.message }`, false)
-		}
-	} finally {
-		isStreaming = false
-		currentStreamController = null
-		sendBtn.disabled = false
-		userInput.focus()
-
-		// Remove cursor if still exists
-		if (cursorSpan.parentNode) {
-			cursorSpan.parentNode.removeChild(cursorSpan)
-		}
-	}
-}
-
-// Handle non-streaming response
-async function handleNonStreamingResponse(userMessage) {
-	const typingIndicator = showTypingIndicator()
-
-	try {
-		// Auto-select first model if none selected
-		if (!currentModel && availableModels.length > 0) {
-			currentModel = availableModels[0].id
-			document.getElementById('modelSelector').value = currentModel
-		}
-
-		const requestBody = {
-			model      : currentModel || 'local-model',
-			messages   : [{ role: 'user', content: userMessage }],
-			stream     : false,
-			temperature: 0.7,
-			max_tokens : 2000
-		}
-
-		const response = await fetch(`${ getApiUrl() }/chat/completions`, {
-			method : 'POST',
-			headers: {
-				'Content-Type' : 'application/json',
-				'Authorization': `Bearer ${ API_KEY }`
-			},
-			body   : JSON.stringify(requestBody)
-		})
-
-		if (!response.ok) {
-			throw new Error(`API Error: ${ response.status } ${ response.statusText }`)
-		}
-
-		const data = await response.json()
-		removeTypingIndicator(typingIndicator)
-
-		if (data.choices && data.choices.length > 0) {
-			const assistantReply = data.choices[0].message.content
-			addMessage('assistant', assistantReply, markdownToggle.checked)
-		} else {
-			addMessage('assistant', 'No response generated.', false)
-		}
-
-	} catch (error) {
-		console.error('Error:', error)
-		removeTypingIndicator(typingIndicator)
-		addMessage('assistant', `Error: ${ error.message }`, false)
-	} finally {
-		sendBtn.disabled = false
-		userInput.focus()
-	}
-}
-
-function loadChatHistory() {
-	chatHistory.forEach(message => {
-		addMessage(message.role, message.content, message.markdown, message.messageId)
-	})
-}
-
-// Stop current stream
-function stopStream() {
-	if (currentStreamController) {
-		currentStreamController.abort()
-		isStreaming = false
-		currentStreamController = null
-		sendBtn.disabled = false
-		sendBtn.textContent = 'Send'
-	}
-}
-
-// Send message
-async function sendMessage() {
-	const userMessage = userInput.value.trim()
-	if (!userMessage || isStreaming) return
-
-	addMessage('user', userMessage)
-	userInput.value = ''
-	userInput.style.height = 'auto'
-	sendBtn.disabled = true
-
-	if (streamToggle.checked) {
-		sendBtn.textContent = 'Stop'
-		handleStreamingResponse(userMessage)
-	} else {
-		handleNonStreamingResponse(userMessage)
-	}
-}
-
-// Event Listeners
-sendBtn.addEventListener('click', () => {
-	if (isStreaming) {
-		stopStream()
-	} else {
-		sendMessage()
-	}
-})
-
-userInput.addEventListener('keydown', (e) => {
-	if (e.key === 'Enter' && !e.shiftKey) {
-		e.preventDefault()
-		if (!isStreaming) {
-			sendMessage()
-		}
-	}
-})
-
-// Initialize on load
-window.onload = () => {
-	userInput.focus()
-	fetchModels()
-
-	// Load chat history
-	loadChatHistory()
-
-	// Add welcome message
-	setTimeout(() => {
-		const welcomeMessage = `# Welcome to LM Studio Chat! 🚀
-
-I now support **real-time streaming responses** and **dark, readable text formatting**.
-
-## Features:
-1. **Streaming Mode** (enabled by default) - Watch responses appear word-by-word
-2. **Markdown Rendering** - Proper formatting for code, lists, tables, and more
-3. **Readable Dark Text** - No more eye strain from light gray text
-4. **Model Detection** - Automatically shows your loaded model
-
-## Try it out:
-- Ask a technical question to see **code formatting**
-- Type \`ls -la\` in a code block
-- Request a **numbered list** or table
-- Watch the streaming response in real-time!
-
-> *Tip: You can toggle streaming and markdown using the checkboxes below.*`
-		addMessage('assistant', welcomeMessage, true)
-	}, 500)
-}
-</script>
+<script src="chatv2.js"></script>
 </body>
 </html>