diff --git a/gpt4all-chat/chatllm.cpp b/gpt4all-chat/chatllm.cpp
index cdb89fe8..4f87d150 100644
--- a/gpt4all-chat/chatllm.cpp
+++ b/gpt4all-chat/chatllm.cpp
@@ -461,8 +461,6 @@ bool ChatLLM::promptInternal(const QList<QString> &collectionList, const QString
     QString instructPrompt = augmentedTemplate.join("\n").arg(prompt);
 
     int n_threads = MySettings::globalInstance()->threadCount();
-    if (n_threads <= 0)
-        n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency());
 
     m_stopGenerating = false;
     auto promptFunc = std::bind(&ChatLLM::handlePrompt, this, std::placeholders::_1);
@@ -773,8 +771,6 @@ void ChatLLM::processSystemPrompt()
     const float repeat_penalty = MySettings::globalInstance()->modelRepeatPenalty(m_modelInfo);
     const int32_t repeat_penalty_tokens = MySettings::globalInstance()->modelRepeatPenaltyTokens(m_modelInfo);
     int n_threads = MySettings::globalInstance()->threadCount();
-    if (n_threads <= 0)
-        n_threads = std::min(4, (int32_t) std::thread::hardware_concurrency());
     m_ctx.n_predict = n_predict;
     m_ctx.top_k = top_k;
     m_ctx.top_p = top_p;
diff --git a/gpt4all-chat/mysettings.cpp b/gpt4all-chat/mysettings.cpp
index 6d40d76a..9a4e17ff 100644
--- a/gpt4all-chat/mysettings.cpp
+++ b/gpt4all-chat/mysettings.cpp
@@ -8,7 +8,7 @@
 #include <QStandardPaths>
 #include <QUrl>
 
-static int      default_threadCount         = 0;
+static int      default_threadCount         = std::min(4, (int32_t) std::thread::hardware_concurrency());
 static bool     default_saveChats           = false;
 static bool     default_saveChatGPTChats    = true;
 static bool     default_serverChat          = false;
@@ -349,7 +349,10 @@ int MySettings::threadCount() const
 {
     QSettings setting;
     setting.sync();
-    return setting.value("threadCount", default_threadCount).toInt();
+    int c = setting.value("threadCount", default_threadCount).toInt();
+    c = std::max(c, 1);
+    c = std::min(c, QThread::idealThreadCount());
+    return c;
 }
 
 void MySettings::setThreadCount(int c)
@@ -357,6 +360,8 @@ void MySettings::setThreadCount(int c)
     if (threadCount() == c)
         return;
 
+    c = std::max(c, 1);
+    c = std::min(c, QThread::idealThreadCount());
     QSettings setting;
     setting.setValue("threadCount", c);
     setting.sync();
diff --git a/gpt4all-chat/qml/ApplicationSettings.qml b/gpt4all-chat/qml/ApplicationSettings.qml
index b0208aac..15333b7c 100644
--- a/gpt4all-chat/qml/ApplicationSettings.qml
+++ b/gpt4all-chat/qml/ApplicationSettings.qml
@@ -101,7 +101,7 @@ MySettingsTab {
         MyTextField {
             text: MySettings.threadCount
             color: theme.textColor
-            ToolTip.text: qsTr("Amount of processing threads to use, a setting of 0 will use the lesser of 4 or your number of CPU threads")
+            ToolTip.text: qsTr("Amount of processing threads to use bounded by 1 and number of logical processors")
             ToolTip.visible: hovered
             Layout.row: 3
             Layout.column: 1