diff --git a/addon/globalPlugins/openai/__init__.py b/addon/globalPlugins/openai/__init__.py
index 84c6d49..ce8bc9b 100644
--- a/addon/globalPlugins/openai/__init__.py
+++ b/addon/globalPlugins/openai/__init__.py
@@ -11,21 +11,19 @@
 import ui
 from logHandler import log
 from scriptHandler import script, getLastScriptRepeatCount
+from . import configspec
+from . import updatecheck
 from .apikeymanager import APIKeyManager
 from .consts import (
 	ADDON_DIR, DATA_DIR,
-	DEFAULT_MODEL, DEFAULT_TOP_P, DEFAULT_N,
-	TOP_P_MIN, TOP_P_MAX,
-	N_MIN, N_MAX,
-	TTS_VOICES, TTS_MODELS, TTS_DEFAULT_VOICE, TTS_DEFAULT_MODEL
+	LIBS_DIR_PY,
+	TTS_MODELS, TTS_VOICES
 )
 from .recordthread import RecordThread
-
-additionalLibsPath = os.path.join(ADDON_DIR, "lib")
-sys.path.insert(0, additionalLibsPath)
+sys.path.insert(0, LIBS_DIR_PY)
 import mss
 from openai import OpenAI
-sys.path.remove(additionalLibsPath)
+sys.path.remove(LIBS_DIR_PY)
 
 addonHandler.initTranslation()
 
@@ -34,40 +32,9 @@
 	ROOT_ADDON_DIR
 ).manifest
 
-confSpecs = {
-	"use_org": "boolean(default=False)",
-	"model": f"string(default={DEFAULT_MODEL.name})",
-	"topP": f"integer(min={TOP_P_MIN}, max={TOP_P_MAX}, default={DEFAULT_TOP_P})",
-	"n": f"integer(min={N_MIN}, max={N_MAX}, default={DEFAULT_N})",
-	"stream": "boolean(default=True)",
-	"TTSModel": f"option({', '.join(TTS_MODELS)}, default={TTS_DEFAULT_MODEL})",
-	"TTSVoice": f"option({', '.join(TTS_VOICES)}, default={TTS_DEFAULT_VOICE})",
-	"blockEscapeKey": "boolean(default=False)",
-	"conversationMode": "boolean(default=True)",
-	"saveSystem": "boolean(default=true)",
-	"advancedMode": "boolean(default=False)",
-	"images": {
-		"maxHeight": "integer(min=0, default=720)",
-		"maxWidth": "integer(min=0, default=0)",
-		"quality": "integer(min=0, max=100, default=85)",
-		"resize": "boolean(default=False)",
-		"resizeInfoDisplayed": "boolean(default=False)",
-		"useCustomPrompt": "boolean(default=False)",
-		"customPromptText": 'string(default="")'
-	},
-	"audio": {
-		"sampleRate": "integer(min=8000, max=48000, default=16000)",
-		"channels": "integer(min=1, max=2, default=1)",
-		"dtype": "string(default=int16)"
-	},
-	"renewClient": "boolean(default=False)",
-	"debug": "boolean(default=False)"
-}
-config.conf.spec["OpenAI"] = confSpecs
-conf = config.conf["OpenAI"]
-
 NO_AUTHENTICATION_KEY_PROVIDED_MSG = _("No authentication key provided. Please set it in the Preferences dialog.")
 
+conf = config.conf["OpenAI"]
 api_key_manager = APIKeyManager(DATA_DIR)
 
 class SettingsDlg(gui.settingsDialogs.SettingsPanel):
@@ -76,6 +43,31 @@ class SettingsDlg(gui.settingsDialogs.SettingsPanel):
 
 	def makeSettings(self, settingsSizer):
 		sHelper = gui.guiHelper.BoxSizerHelper(self, sizer=settingsSizer)
+
+		updateGroupLabel = _("Update")
+		updateSizer = wx.StaticBoxSizer(wx.VERTICAL, self, label=updateGroupLabel)
+		updateBox = updateSizer.GetStaticBox()
+		updateGroup = gui.guiHelper.BoxSizerHelper(self, sizer=updateSizer)
+
+		self.updateCheck = updateGroup.addItem(
+			wx.CheckBox(
+				updateBox,
+				label=_("Check for &updates on startup and periodically")
+			)
+		)
+		self.updateCheck.SetValue(conf["update"]["check"])
+
+		self.updateChannel = updateGroup.addLabeledControl(
+			_("&Channel:"),
+			wx.Choice,
+			choices=["stable", "dev"]
+		)
+		self.updateChannel.SetSelection(
+			1 if conf["update"]["channel"] == "dev" else 0
+		)
+
+		sHelper.addItem(updateSizer)
+
 		APIKey = api_key_manager.get_api_key()
 		if not APIKey: APIKey = ''
 		APIKeyOrg = api_key_manager.get_api_key(use_org=True)
@@ -224,7 +216,7 @@ def makeSettings(self, settingsSizer):
 
 		self.useCustomPrompt = imageGroup.addItem(
 			wx.CheckBox(
-				imageBox, 
+				imageBox,
 				label=_("Customize default text &prompt")
 			)
 		)
@@ -265,6 +257,8 @@ def onDefaultPrompt(self, evt):
 			self.customPromptText.Enable(False)
 
 	def onSave(self):
+		conf["update"]["check"] = self.updateCheck.GetValue()
+		conf["update"]["channel"] = self.updateChannel.GetString(self.updateChannel.GetSelection())
 		api_key = self.APIKey.GetValue().strip()
 		api_key_manager.save_api_key(api_key)
 		api_key_org = self.org_key.GetValue().strip()
@@ -326,6 +320,9 @@ def createMenu(self):
 			_("Show the Open AI dialog")
 		)
 		gui.mainFrame.sysTrayIcon.Bind(wx.EVT_MENU, self.onShowMainDialog, item)
+
+		self.submenu.AppendSeparator()
+
 		item = self.submenu.Append(
 			wx.ID_ANY,
 			_("API &keys"),
@@ -345,6 +342,19 @@ def createMenu(self):
 		)
 		gui.mainFrame.sysTrayIcon.Bind(wx.EVT_MENU, self.onGitRepo, item)
 
+		self.submenu.AppendSeparator()
+
+		item = self.submenu.Append(
+			wx.ID_ANY,
+			_("Check for &updates..."),
+			_("Check for updates")
+		)
+		gui.mainFrame.sysTrayIcon.Bind(
+			wx.EVT_MENU,
+			self.onCheckForUpdates,
+			item
+		)
+
 		addon_name = ADDON_INFO["name"]
 		addon_version = ADDON_INFO["version"]
 		self.submenu_item = gui.mainFrame.sysTrayIcon.menu.InsertMenu(
@@ -381,6 +391,12 @@ def onDocumentation(self, evt):
 				os.startfile(fp)
 				break
 
+	def onCheckForUpdates(self, evt):
+		updatecheck.check_update(
+			auto=False
+		)
+		updatecheck.update_last_check()
+
 	def terminate(self):
 		gui.settingsDialogs.NVDASettingsDialog.categoryClasses.remove(SettingsDlg)
 		gui.mainFrame.sysTrayIcon.menu.DestroyItem(self.submenu_item)
diff --git a/addon/globalPlugins/openai/configspec.py b/addon/globalPlugins/openai/configspec.py
new file mode 100644
index 0000000..5c1abd8
--- /dev/null
+++ b/addon/globalPlugins/openai/configspec.py
@@ -0,0 +1,49 @@
+import config
+from .consts import (
+	DEFAULT_MODEL,
+	DEFAULT_TOP_P,
+	DEFAULT_N,
+	TOP_P_MIN,
+	TOP_P_MAX,
+	N_MIN,
+	N_MAX,
+	TTS_MODELS,
+	TTS_DEFAULT_MODEL,
+	TTS_VOICES,
+	TTS_DEFAULT_VOICE
+)
+
+confSpecs = {
+	"update": {
+		"check": "boolean(default=True)",
+		"channel": "string(default='stable')"
+	},
+	"use_org": "boolean(default=False)",
+	"model": f"string(default={DEFAULT_MODEL.name})",
+	"topP": f"integer(min={TOP_P_MIN}, max={TOP_P_MAX}, default={DEFAULT_TOP_P})",
+	"n": f"integer(min={N_MIN}, max={N_MAX}, default={DEFAULT_N})",
+	"stream": "boolean(default=True)",
+	"TTSModel": f"option({', '.join(TTS_MODELS)}, default={TTS_DEFAULT_MODEL})",
+	"TTSVoice": f"option({', '.join(TTS_VOICES)}, default={TTS_DEFAULT_VOICE})",
+	"blockEscapeKey": "boolean(default=False)",
+	"conversationMode": "boolean(default=True)",
+	"saveSystem": "boolean(default=true)",
+	"advancedMode": "boolean(default=False)",
+	"images": {
+		"maxHeight": "integer(min=0, default=720)",
+		"maxWidth": "integer(min=0, default=0)",
+		"quality": "integer(min=0, max=100, default=85)",
+		"resize": "boolean(default=False)",
+		"resizeInfoDisplayed": "boolean(default=False)",
+		"useCustomPrompt": "boolean(default=False)",
+		"customPromptText": 'string(default="")'
+	},
+	"audio": {
+		"sampleRate": "integer(min=8000, max=48000, default=16000)",
+		"channels": "integer(min=1, max=2, default=1)",
+		"dtype": "string(default=int16)"
+	},
+	"renewClient": "boolean(default=False)",
+	"debug": "boolean(default=False)"
+}
+config.conf.spec["OpenAI"] = confSpecs
diff --git a/addon/globalPlugins/openai/consts.py b/addon/globalPlugins/openai/consts.py
index bdeedd8..64bae4b 100644
--- a/addon/globalPlugins/openai/consts.py
+++ b/addon/globalPlugins/openai/consts.py
@@ -1,4 +1,5 @@
 import os
+import sys
 import globalVars
 import addonHandler
 from .model import Model
@@ -54,4 +55,13 @@
 	"organization of the interface. If the image does not correspond to a computer screen, just generate "
 	"a detailed visual description. If the user sends an image alone without additional instructions in text, "
 	"describe the image exactly as prescribed in this system prompt. Adhere strictly to the instructions in "
-	"this system prompt to describe images. Don’t add any additional details unless the user specifically ask you.")
+	"this system prompt to describe images. Don’t add any additional details unless the user specifically ask you."
+)
+LIBS_DIR = os.path.join(DATA_DIR, "libs")
+LIBS_DIR_PY = os.path.join(
+	LIBS_DIR,
+	"lib_py%s.%s" % (
+		sys.version_info.major,
+		sys.version_info.minor
+	)
+)
diff --git a/addon/globalPlugins/openai/imagehelper.py b/addon/globalPlugins/openai/imagehelper.py
index 02476cc..c14ecbd 100644
--- a/addon/globalPlugins/openai/imagehelper.py
+++ b/addon/globalPlugins/openai/imagehelper.py
@@ -2,14 +2,14 @@
 import os
 import sys
 from logHandler import log
-from .consts import ADDON_DIR
+from .consts import ADDON_DIR, LIBS_DIR_PY
 
-additionalLibsPath = os.path.join(ADDON_DIR, "lib")
-sys.path.insert(0, additionalLibsPath)
+sys.path.append(LIBS_DIR_PY)
 from openai import OpenAI
 from PIL import Image
 import fractions
-sys.path.remove(additionalLibsPath)
+sys.path.remove(LIBS_DIR_PY)
+
 
 def get_image_dimensions(path):
 	"""
diff --git a/addon/globalPlugins/openai/maindialog.py b/addon/globalPlugins/openai/maindialog.py
index 4feee46..21cf516 100644
--- a/addon/globalPlugins/openai/maindialog.py
+++ b/addon/globalPlugins/openai/maindialog.py
@@ -19,6 +19,7 @@
 from logHandler import log
 from .consts import (
 	ADDON_DIR, DATA_DIR,
+	LIBS_DIR_PY,
 	MODELS, MODEL_VISION,
 	TOP_P_MIN, TOP_P_MAX,
 	N_MIN, N_MAX,
@@ -33,11 +34,10 @@
 from .recordthread import RecordThread
 from .resultevent import ResultEvent, EVT_RESULT_ID
 
-additionalLibsPath = os.path.join(ADDON_DIR, "lib")
-sys.path.insert(0, additionalLibsPath)
+sys.path.insert(0, LIBS_DIR_PY)
 import openai
 import markdown2
-sys.path.remove(additionalLibsPath)
+sys.path.remove(LIBS_DIR_PY)
 
 addonHandler.initTranslation()
 
@@ -771,12 +771,19 @@ def onModelChange(self, evt):
 			defaultMaxOutputToken = 1024
 		self.maxTokens.SetValue(defaultMaxOutputToken)
 		if self.conf["advancedMode"]:
-			self.temperature.SetRange(0, model.maxTemperature * 100)
+			self.temperature.SetRange(
+				0,
+				int(model.maxTemperature * 100)
+			)
 			key_temperature = "temperature_%s" % model.name
 			if key_temperature in self.data:
-				self.temperature.SetValue(self.data[key_temperature])
+				self.temperature.SetValue(
+					int(self.data[key_temperature])
+				)
 			else:
-				self.temperature.SetValue(model.defaultTemperature * 100) 
+				self.temperature.SetValue(
+					int(model.defaultTemperature * 100)
+				)
 
 	def onOk(self, evt):
 		if not self.promptText.GetValue().strip() and not self.pathList:
diff --git a/addon/globalPlugins/openai/recordthread.py b/addon/globalPlugins/openai/recordthread.py
index c4657cb..edd34e7 100644
--- a/addon/globalPlugins/openai/recordthread.py
+++ b/addon/globalPlugins/openai/recordthread.py
@@ -16,14 +16,13 @@
 import core
 import ui
 
-from .consts import ADDON_DIR, DATA_DIR
+from .consts import ADDON_DIR, DATA_DIR, LIBS_DIR_PY
 from .resultevent import ResultEvent
 
-additionalLibsPath = os.path.join(ADDON_DIR, "lib")
-sys.path.insert(0, additionalLibsPath)
+sys.path.insert(0, LIBS_DIR_PY)
 import numpy as np
 import sounddevice as sd
-sys.path.remove(additionalLibsPath)
+sys.path.remove(LIBS_DIR_PY)
 
 addonHandler.initTranslation()
 
diff --git a/addon/globalPlugins/openai/updatecheck.py b/addon/globalPlugins/openai/updatecheck.py
new file mode 100644
index 0000000..2b1f2a5
--- /dev/null
+++ b/addon/globalPlugins/openai/updatecheck.py
@@ -0,0 +1,225 @@
+import hashlib
+import json
+import os
+import shutil
+import sys
+import time
+import urllib.request
+import zipfile
+import gui
+import wx
+
+import addonHandler
+import config
+import core
+import ui
+import versionInfo
+from logHandler import log
+
+from .consts import ADDON_DIR, DATA_DIR, LIBS_DIR, LIBS_DIR_PY
+
+addonHandler.initTranslation()
+
+ROOT_ADDON_DIR = "\\".join(ADDON_DIR.split(os.sep)[:-2])
+ADDON_INFO = addonHandler.Addon(ROOT_ADDON_DIR).manifest
+LAST_CHECK_FP = os.path.join(DATA_DIR, "last_check")
+LIB_REV_FP = os.path.join(LIBS_DIR_PY, "libs_rev.txt")
+URL_LATEST_RELEASE = "https://andreabc.net/projects/NVDA_addons/OpenAI/version.json"
+
+NVDA_VERSION = f"{versionInfo.version_year}.{versionInfo.version_major}.{versionInfo.version_minor}"
+PYTHON_VERSION = f"{sys.version_info.major}.{sys.version_info.minor}"
+
+conf = config.conf["OpenAI"]["update"]
+
+
+def ensure_dir_exists(directory: str):
+	"""Ensure that the specified directory exists, creating it if necessary."""
+	if not os.path.exists(directory):
+		os.mkdir(directory)
+
+ensure_dir_exists(DATA_DIR)
+
+_lib_rev = ""
+
+def get_lib_rev():
+	"""Return the current revision of the OpenAI dependencies."""
+	global _lib_rev
+	if _lib_rev:
+		return _lib_rev
+	if os.path.exists(LIB_REV_FP):
+		try:
+			with open(LIB_REV_FP) as file:
+				_lib_rev = file.read().strip()
+		except ValueError:
+			pass
+	return _lib_rev
+
+
+_last_check = 0
+
+def get_last_check():
+	"""Return the timestamp for the latest update check"""
+	global _last_check
+	if _last_check:
+		return _last_check
+	if os.path.exists(LAST_CHECK_FP):
+		try:
+			with open(LAST_CHECK_FP) as file:
+				_last_check = float(file.read().strip())
+		except ValueError:
+			pass
+	return _last_check
+
+
+def update_last_check():
+	"""Update the timestamp for the last check in the configuration."""
+	global _last_check
+	_last_check = time.time()
+	with open(LAST_CHECK_FP, "w") as file:
+		file.write(str(_last_check))
+
+
+def check_addon_version(data: dict, auto: bool):
+	"""Check if the addon is up to date and notify the user accordingly."""
+	local_version = ADDON_INFO["version"]
+	remote_version = data["addon_version"]
+	if local_version.split('-')[0] != remote_version:
+		# Translators: This is the message displayed when a new version of the add-on is available.
+		msg = _(
+			"New version available: %s. Your version is %s. "
+			"You can update from the add-on store "
+			"or from the GitHub repository."
+		) % (
+			remote_version,
+			local_version
+		)
+		gui.messageBox(
+			msg,
+			# Translators: This is the title of the message displayed when a new version of the add-on is available.
+			_("OpenAI update"),
+			wx.OK | wx.ICON_INFORMATION
+		)
+	elif not auto:
+		# Translators: This is the message displayed when the user checks for updates manually and there are no updates available.
+		msg = _("You have the latest version of OpenAI add-on installed.")
+		gui.messageBox(
+			msg,
+			# Translators: This is the title of the message displayed when the user checks for updates manually and there are no updates available.
+			_("OpenAI update"),
+			wx.OK | wx.ICON_INFORMATION
+		)
+
+
+def load_remote_data(auto: bool):
+	"""Load and return the remote version and dependency information."""
+	channel = conf["channel"]
+	params = {
+		"nvda_version": NVDA_VERSION,
+		"python_version": PYTHON_VERSION,
+		"addon_version": ADDON_INFO["version"],
+		"addon_channel": channel,
+		"lib_rev": get_lib_rev()
+	}
+	request_url = f"{URL_LATEST_RELEASE}?{urllib.parse.urlencode(params)}"
+	with urllib.request.urlopen(request_url) as response:
+		return json.loads(response.read())
+
+
+def handle_data_update(response_data: dict, auto: bool):
+	"""Check the addon and dependencies version and handle updates if necessary."""
+	check_addon_version(response_data, auto)
+	if (get_lib_rev() != response_data["libs_rev"]) or not os.path.exists(LIBS_DIR):
+		if offer_data_update(response_data):
+			update_dependency_files(response_data)
+	else:
+		update_last_check()
+
+def check_file_hash(file_path: str, expected_hash: str) -> bool:
+	"""Check that the SHA256 hash of the file at file_path matches expected_hash."""
+	with open(file_path, "rb") as file:
+		file_hash = hashlib.sha256(file.read()).hexdigest()
+	return file_hash == expected_hash
+
+def offer_data_update(data: dict) -> bool:
+	"""Offer to update the OpenAI dependencies and return the user's choice."""
+	# Translators: This is the message displayed when a new version of the OpenAI dependencies is available.
+	msg = _("New OpenAI dependencies revision available: %s. Your version is %s. Update now?") % (
+		data["libs_rev"],
+		get_lib_rev() or _("unknown")
+	)
+
+	result = gui.messageBox(
+		msg,
+		# Translators: This is the title of the message displayed when a new version of the OpenAI dependencies is available.
+		_("OpenAI dependencies update"),
+		wx.YES_NO | wx.ICON_QUESTION
+	)
+	return result == wx.YES
+
+
+def update_dependency_files(data: dict):
+	"""Handle the downloading and extraction of updated dependencies."""
+	ui.message(
+		# Translators: This is the message emitted when updating the OpenAI dependencies.
+		_("Updating OpenAI dependencies... Please wait")
+	)
+	try:
+		with urllib.request.urlopen(data["libs_download_url"]) as response:
+			zip_file_content = response.read()
+
+		zip_path = os.path.join(DATA_DIR, "libs.zip")
+		with open(zip_path, "wb") as file:
+			file.write(zip_file_content)
+
+		if not check_file_hash(zip_path, data["libs_hash"]):
+			raise ValueError("Libs hash mismatch")
+
+		if os.path.exists(LIBS_DIR_PY):
+			shutil.rmtree(LIBS_DIR_PY)
+
+		with zipfile.ZipFile(zip_path, "r") as zip_file:
+			zip_file.extractall(LIBS_DIR)
+
+		os.remove(zip_path)
+		update_last_check()
+		gui.messageBox(
+			# Translators: This is the message displayed when the OpenAI dependencies are updated successfully.
+			_("Dependencies updated successfully. NVDA will now restart to apply the changes."),
+			# Translators: This is the title of the message displayed when the OpenAI dependencies are updated successfully.
+			_("Success"),
+			wx.OK | wx.ICON_INFORMATION
+		)
+		core.restart()
+
+	except (
+		urllib.error.URLError,
+		zipfile.BadZipFile,
+		ValueError,
+		OSError
+	) as e:
+		log.error(f"Error updating OpenAI dependencies: {e}")
+		gui.messageBox(
+			# Translators: This is the message displayed when the OpenAI dependencies cannot be updated.
+			_("Error updating dependencies. Please restart NVDA and try again."),
+			# Translators: This is the title of the message displayed when the OpenAI dependencies cannot be updated.
+			_("Error"),
+			wx.OK | wx.ICON_ERROR
+		)
+
+
+def check_update(auto: bool = True):
+	"""Check for updates to OpenAI addon, including new versions and dependencies."""
+	log.info("Checking for Open AI updates (auto=%s)", auto)
+	try:
+		data = load_remote_data(auto)
+		wx.CallAfter(handle_data_update, data, auto)
+	except urllib.error.URLError as e:
+		log.error(f"Error checking Open AI update: {e}")
+
+
+if (
+	(conf["check"] and get_last_check() + 86400 * 3 < time.time())
+	or not os.path.exists(LIBS_DIR)
+	or not get_lib_rev()
+):
+	check_update()
diff --git a/buildVars.py b/buildVars.py
index e996fe7..e148498 100644
--- a/buildVars.py
+++ b/buildVars.py
@@ -54,7 +54,7 @@ def _(arg):
 	# Minimum NVDA version supported (e.g. "2018.3.0", minor version is optional)
 	"addon_minimumNVDAVersion": "2023.1",
 	# Last NVDA version supported/tested (e.g. "2018.4.0", ideally more recent than minimum version)
-	"addon_lastTestedNVDAVersion": "2023.3",
+	"addon_lastTestedNVDAVersion": "2024.1",
 	# Add-on update channel (default is None, denoting stable releases,
 	# and for development releases, use "dev".)
 	# Do not change unless you know what you are doing!