diff --git a/gpt4all-chat/chatllm.cpp b/gpt4all-chat/chatllm.cpp
index 088083d5..0f5da45f 100644
--- a/gpt4all-chat/chatllm.cpp
+++ b/gpt4all-chat/chatllm.cpp
@@ -788,13 +788,18 @@ void ChatLLM::processSystemPrompt()
     if (!isModelLoaded() || m_processedSystemPrompt || m_isServer)
         return;
 
+    const std::string systemPrompt = MySettings::globalInstance()->modelSystemPrompt(m_modelInfo).toStdString();
+    if (systemPrompt.empty()) {
+        m_processedSystemPrompt = true;
+        return;
+    }
+
     m_stopGenerating = false;
     auto promptFunc = std::bind(&ChatLLM::handleSystemPrompt, this, std::placeholders::_1);
     auto responseFunc = std::bind(&ChatLLM::handleSystemResponse, this, std::placeholders::_1,
         std::placeholders::_2);
     auto recalcFunc = std::bind(&ChatLLM::handleSystemRecalculate, this, std::placeholders::_1);
 
-    const std::string systemPrompt = MySettings::globalInstance()->modelSystemPrompt(m_modelInfo).toStdString();
     const int32_t n_predict = MySettings::globalInstance()->modelMaxLength(m_modelInfo);
     const int32_t top_k = MySettings::globalInstance()->modelTopK(m_modelInfo);
     const float top_p = MySettings::globalInstance()->modelTopP(m_modelInfo);
diff --git a/gpt4all-chat/metadata/models.json b/gpt4all-chat/metadata/models.json
index f0578f71..13928a62 100644
--- a/gpt4all-chat/metadata/models.json
+++ b/gpt4all-chat/metadata/models.json
@@ -9,6 +9,7 @@
     "parameters": "13 billion",
     "quant": "q4_0",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Best overall model</strong><br><ul><li>Instruction based<li>Gives very long responses<li>Finetuned with only 1k of high-quality data<li>Trained by Microsoft and Peking University<li>Cannot be used commercially</ul"
   },
   {
@@ -22,6 +23,7 @@
     "parameters": "7 billion",
     "quant": "q4_0",
     "type": "Falcon",
+    "systemPrompt": " ",
     "description": "<strong>Best overall smaller model</strong><br><ul><li>Fast responses</li><li>Instruction based</li><li>Trained by TII<li>Finetuned by Nomic AI<li>Licensed for commercial use</ul>",
     "url": "https://huggingface.co/nomic-ai/gpt4all-falcon-ggml/resolve/main/ggml-model-gpt4all-falcon-q4_0.bin",
     "promptTemplate": "### Instruction:\n%1\n### Response:\n"
@@ -37,6 +39,7 @@
     "parameters": "13 billion",
     "quant": "q4_0",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Extremely good model</strong><br><ul><li>Instruction based<li>Gives long responses<li>Curated with 300,000 uncensored instructions<li>Trained by Nous Research<li>Cannot be used commercially</ul>",
     "url": "https://huggingface.co/TheBloke/Nous-Hermes-13B-GGML/resolve/main/nous-hermes-13b.ggmlv3.q4_0.bin",
     "promptTemplate": "### Instruction:\n%1\n### Response:\n"
@@ -51,6 +54,7 @@
     "parameters": "7 billion",
     "quant": "q4_0",
     "type": "GPT-J",
+    "systemPrompt": " ",
     "description": "<strong>Creative model can be used for commercial purposes</strong><br><ul><li>Fast responses<li>Creative responses</li><li>Instruction based</li><li>Trained by Nomic AI<li>Licensed for commercial use</ul>"
   },
   {
@@ -64,6 +68,7 @@
     "parameters": "13 billion",
     "quant": "q4_0",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Very good overall model</strong><br><ul><li>Instruction based<li>Based on the same dataset as Groovy<li>Slower than Groovy, with higher quality responses<li>Trained by Nomic AI<li>Cannot be used commercially</ul>",
     "url": "https://huggingface.co/TheBloke/GPT4All-13B-snoozy-GGML/resolve/main/GPT4All-13B-snoozy.ggmlv3.q4_0.bin"
   },
@@ -140,6 +145,7 @@
     "parameters": "7 billion",
     "quant": "q4_2",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Good small model - trained by teams from UC Berkeley, CMU, Stanford, MBZUAI, and UC San Diego</strong><br><ul><li>Instruction based<li>Cannot be used commercially</ul>"
   },
   {
@@ -152,6 +158,7 @@
     "parameters": "13 billion",
     "quant": "q4_2",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Good larger model - trained by teams from UC Berkeley, CMU, Stanford, MBZUAI, and UC San Diego</strong><br><ul><li>Instruction based<li>Cannot be used commercially</ul>"
   },
   {
@@ -164,6 +171,7 @@
     "parameters": "7 billion",
     "quant": "q4_2",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Good small model - trained by by Microsoft and Peking University</strong><br><ul><li>Instruction based<li>Cannot be used commercially</ul>"
   },
   {
@@ -190,6 +198,7 @@
     "parameters": "7 billion",
     "quant": "q4_0",
     "type": "MPT",
+    "systemPrompt": " ",
     "description": "<strong>Mosaic's instruction model</strong><br><ul><li>Instruction based<li>Trained by Mosaic ML<li>Licensed for commercial use</ul>"
   },
   {
@@ -203,6 +212,7 @@
     "parameters": "7 billion",
     "quant": "q4_0",
     "type": "MPT",
+    "systemPrompt": " ",
     "description": "<strong>Trained for text completion with no assistant finetuning</strong><br><ul><li>Completion based<li>Trained by Mosaic ML<li>Licensed for commercial use</ul>"
   },
   {
@@ -215,6 +225,7 @@
     "parameters": "13 billion",
     "quant": "q4_0",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Trained on ~180,000 instructions</strong><br><ul><li>Instruction based<li>Trained by Nous Research<li>Cannot be used commercially</ul>"
   },
   {
@@ -228,6 +239,7 @@
     "parameters": "13 billion",
     "quant": "q4_0",
     "type": "LLaMA",
+    "systemPrompt": " ",
     "description": "<strong>Trained on uncensored assistant data and instruction data</strong><br><ul><li>Instruction based<li>Cannot be used commercially</ul>",
     "url": "https://huggingface.co/TheBloke/WizardLM-13B-Uncensored-GGML/resolve/main/wizardLM-13B-Uncensored.ggmlv3.q4_0.bin"
   },
@@ -243,6 +255,7 @@
     "parameters": "3 billion",
     "quant": "f16",
     "type": "Replit",
+    "systemPrompt": " ",
     "description": "<strong>Trained on subset of the Stack</strong><br><ul><li>Code completion based<li>Licensed for commercial use</ul>",
     "url": "https://huggingface.co/nomic-ai/ggml-replit-code-v1-3b/resolve/main/ggml-replit-code-v1-3b.bin"
   }