diff --git a/extensions/inference-groq-extension/resources/models.json b/extensions/inference-groq-extension/resources/models.json
index 81275f47c..6fce1c71b 100644
--- a/extensions/inference-groq-extension/resources/models.json
+++ b/extensions/inference-groq-extension/resources/models.json
@@ -69,9 +69,9 @@
     ],
     "id": "gemma-7b-it",
     "object": "model",
-    "name": "Groq Gemma 7b Instruct",
-    "version": "1.1",
-    "description": "Groq Gemma 7b Instruct with supercharged speed!",
+    "name": "Groq Gemma 7B Instruct",
+    "version": "1.2",
+    "description": "Groq Gemma 7B Instruct with supercharged speed!",
     "format": "api",
     "settings": {},
     "parameters": {
@@ -99,9 +99,9 @@
     ],
     "id": "mixtral-8x7b-32768",
     "object": "model",
-    "name": "Groq Mixtral 8x7b Instruct",
-    "version": "1.1",
-    "description": "Groq Mixtral 8x7b Instruct is Mixtral with supercharged speed!",
+    "name": "Groq Mixtral 8x7B Instruct",
+    "version": "1.2",
+    "description": "Groq Mixtral 8x7B Instruct is Mixtral with supercharged speed!",
     "format": "api",
     "settings": {},
     "parameters": {
diff --git a/extensions/inference-nitro-extension/package.json b/extensions/inference-nitro-extension/package.json
index 1e3ea6d38..425e4b49c 100644
--- a/extensions/inference-nitro-extension/package.json
+++ b/extensions/inference-nitro-extension/package.json
@@ -1,7 +1,7 @@
 {
   "name": "@janhq/inference-cortex-extension",
   "productName": "Cortex Inference Engine",
-  "version": "1.0.15",
+  "version": "1.0.16",
   "description": "This extension embeds cortex.cpp, a lightweight inference engine written in C++. See https://jan.ai.\nAdditional dependencies could be installed to run without Cuda Toolkit installation.",
   "main": "dist/index.js",
   "node": "dist/node/index.cjs.js",
diff --git a/extensions/inference-nitro-extension/resources/models/aya-23-8b/model.json b/extensions/inference-nitro-extension/resources/models/aya-23-8b/model.json
index b82cf2f39..163373014 100644
--- a/extensions/inference-nitro-extension/resources/models/aya-23-8b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/aya-23-8b/model.json
@@ -8,7 +8,7 @@
     "id": "aya-23-8b",
     "object": "model",
     "name": "Aya 23 8B Q4",
-    "version": "1.1",
+    "version": "1.2",
     "description": "Aya 23 can talk upto 23 languages fluently.",
     "format": "gguf",
     "settings": {
@@ -28,7 +28,7 @@
     },
     "metadata": {
       "author": "CohereForAI",
-      "tags": ["7B", "Finetuned","Featured"],
+      "tags": ["7B", "Finetuned"],
       "size": 5056982144
     },
     "engine": "nitro"
diff --git a/extensions/inference-nitro-extension/resources/models/deepseek-coder-1.3b/model.json b/extensions/inference-nitro-extension/resources/models/deepseek-coder-1.3b/model.json
index 53f7f43e9..36fceaad2 100644
--- a/extensions/inference-nitro-extension/resources/models/deepseek-coder-1.3b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/deepseek-coder-1.3b/model.json
@@ -7,8 +7,8 @@
   ],
   "id": "deepseek-coder-1.3b",
   "object": "model",
-  "name": "Deepseek Coder 1.3B Q8",
-  "version": "1.2",
+  "name": "Deepseek Coder 1.3B Instruct Q8",
+  "version": "1.3",
   "description": "Deepseek Coder excelled in project-level code completion with advanced capabilities across multiple programming languages.",
   "format": "gguf",
   "settings": {
diff --git a/extensions/inference-nitro-extension/resources/models/deepseek-coder-34b/model.json b/extensions/inference-nitro-extension/resources/models/deepseek-coder-34b/model.json
index 0a3e58b48..103c4cbcb 100644
--- a/extensions/inference-nitro-extension/resources/models/deepseek-coder-34b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/deepseek-coder-34b/model.json
@@ -7,8 +7,8 @@
   ],
   "id": "deepseek-coder-34b",
   "object": "model",
-  "name": "Deepseek Coder 33B Q4",
-  "version": "1.2",
+  "name": "Deepseek Coder 33B Instruct Q4",
+  "version": "1.3",
   "description": "Deepseek Coder excelled in project-level code completion with advanced capabilities across multiple programming languages.",
   "format": "gguf",
   "settings": {
diff --git a/extensions/inference-nitro-extension/resources/models/gemma-1.1-7b/model.json b/extensions/inference-nitro-extension/resources/models/gemma-1.1-7b/model.json
index a532c1dc3..b29043483 100644
--- a/extensions/inference-nitro-extension/resources/models/gemma-1.1-7b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/gemma-1.1-7b/model.json
@@ -8,7 +8,7 @@
   "id": "gemma-1.1-7b-it",
   "object": "model",
   "name": "Gemma 1.1 7B Q4",
-  "version": "1.2",
+  "version": "1.3",
   "description": "Google's Gemma is built for multilingual purpose",
   "format": "gguf",
   "settings": {
@@ -28,7 +28,7 @@
   },
   "metadata": {
     "author": "Google",
-    "tags": ["7B", "Finetuned", "Featured"],
+    "tags": ["7B", "Finetuned"],
     "size": 5330000000
   },
   "engine": "nitro"
diff --git a/extensions/inference-nitro-extension/resources/models/llama3-8b-instruct/model.json b/extensions/inference-nitro-extension/resources/models/llama3-8b-instruct/model.json
index ced7e1ca8..4d84b9967 100644
--- a/extensions/inference-nitro-extension/resources/models/llama3-8b-instruct/model.json
+++ b/extensions/inference-nitro-extension/resources/models/llama3-8b-instruct/model.json
@@ -7,8 +7,8 @@
     ],
     "id": "llama3-8b-instruct",
     "object": "model",
-    "name": "Llama 3 8B Q4",
-    "version": "1.2",
+    "name": "Llama 3 8B Instruct Q4",
+    "version": "1.4",
     "description": "Meta's Llama 3 excels at general usage situations, including chat, general world knowledge, and coding.",
     "format": "gguf",
     "settings": {
@@ -28,7 +28,7 @@
     },
     "metadata": {
       "author": "MetaAI",
-      "tags": ["8B", "Featured"],
+      "tags": ["8B"],
       "size": 4920000000
     },
     "engine": "nitro"
diff --git a/extensions/inference-nitro-extension/resources/models/llama3.1-70b-instruct/model.json b/extensions/inference-nitro-extension/resources/models/llama3.1-70b-instruct/model.json
index 4d8eab7e3..780ef8465 100644
--- a/extensions/inference-nitro-extension/resources/models/llama3.1-70b-instruct/model.json
+++ b/extensions/inference-nitro-extension/resources/models/llama3.1-70b-instruct/model.json
@@ -7,8 +7,8 @@
   ],
   "id": "llama3.1-70b-instruct",
   "object": "model",
-  "name": "Llama 3.1 70B Q4 Instruct",
-  "version": "1.0",
+  "name": "Llama 3.1 70B Instruct Q4",
+  "version": "1.1",
   "description": "Meta's Llama 3.1 excels at general usage situations, including chat, general world knowledge, and coding.",
   "format": "gguf",
   "settings": {
diff --git a/extensions/inference-nitro-extension/resources/models/llama3.1-8b-instruct/model.json b/extensions/inference-nitro-extension/resources/models/llama3.1-8b-instruct/model.json
index fe44b0b1c..9de8555be 100644
--- a/extensions/inference-nitro-extension/resources/models/llama3.1-8b-instruct/model.json
+++ b/extensions/inference-nitro-extension/resources/models/llama3.1-8b-instruct/model.json
@@ -7,8 +7,8 @@
   ],
   "id": "llama3.1-8b-instruct",
   "object": "model",
-  "name": "Llama 3.1 8B Q4 Instruct",
-  "version": "1.0",
+  "name": "Llama 3.1 8B Instruct Q4",
+  "version": "1.1",
   "description": "Meta's Llama 3.1 excels at general usage situations, including chat, general world knowledge, and coding.",
   "format": "gguf",
   "settings": {
diff --git a/extensions/inference-nitro-extension/resources/models/mistral-ins-7b-q4/model.json b/extensions/inference-nitro-extension/resources/models/mistral-ins-7b-q4/model.json
index d223306f8..88f701466 100644
--- a/extensions/inference-nitro-extension/resources/models/mistral-ins-7b-q4/model.json
+++ b/extensions/inference-nitro-extension/resources/models/mistral-ins-7b-q4/model.json
@@ -7,9 +7,9 @@
   ],
   "id": "mistral-ins-7b-q4",
   "object": "model",
-  "name": "Mistral Instruct 7B Q4",
-  "version": "1.3",
-  "description": "Mistral Instruct 7b model, specifically designed for a comprehensive understanding of the world.",
+  "name": "Mistral 7B Instruct Q4",
+  "version": "1.5",
+  "description": "Mistral 7B Instruct model, specifically designed for a comprehensive understanding of the world.",
   "format": "gguf",
   "settings": {
     "ctx_len": 32768,
@@ -28,7 +28,7 @@
   },
   "metadata": {
     "author": "MistralAI",
-    "tags": ["Featured", "7B", "Foundational Model"],
+    "tags": ["7B", "Foundational Model"],
     "size": 4370000000,
     "cover": "https://raw.githubusercontent.com/janhq/jan/dev/models/mistral-ins-7b-q4/cover.png"
   },
diff --git a/extensions/inference-nitro-extension/resources/models/phi3-3.8b/model.json b/extensions/inference-nitro-extension/resources/models/phi3-3.8b/model.json
index 2a572db92..6459b049d 100644
--- a/extensions/inference-nitro-extension/resources/models/phi3-3.8b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/phi3-3.8b/model.json
@@ -1,20 +1,20 @@
 {
     "sources": [
       {
-        "url": "https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-gguf/resolve/main/Phi-3-mini-4k-instruct-q4.gguf",
-        "filename": "Phi-3-mini-4k-instruct-q4.gguf"
+        "url": "https://huggingface.co/cortexso/phi3/resolve/main/model.gguf",
+        "filename": "model.gguf"
       }
     ],
     "id": "phi3-3.8b",
     "object": "model",
-    "name": "Phi-3 Mini",
-    "version": "1.2",
+    "name": "Phi-3 Mini Instruct Q4",
+    "version": "1.3",
     "description": "Phi-3 Mini is Microsoft's newest, compact model designed for mobile use.",
     "format": "gguf",
     "settings": {
       "ctx_len": 4096,
       "prompt_template": "<|user|>\n{prompt}<|end|>\n<|assistant|>\n",
-      "llama_model_path": "Phi-3-mini-4k-instruct-q4.gguf",
+      "llama_model_path": "model.gguf",
       "ngl": 33
     },
     "parameters": {
@@ -35,4 +35,4 @@
       "size": 2320000000
     },
     "engine": "nitro"
-  }
\ No newline at end of file
+}
\ No newline at end of file
diff --git a/extensions/inference-nitro-extension/resources/models/phi3-medium/model.json b/extensions/inference-nitro-extension/resources/models/phi3-medium/model.json
index ac83ca077..50944b9fe 100644
--- a/extensions/inference-nitro-extension/resources/models/phi3-medium/model.json
+++ b/extensions/inference-nitro-extension/resources/models/phi3-medium/model.json
@@ -7,13 +7,13 @@
     ],
     "id": "phi3-medium",
     "object": "model",
-    "name": "Phi-3 Medium",
-    "version": "1.2",
+    "name": "Phi-3 Medium Instruct Q4",
+    "version": "1.3",
     "description": "Phi-3 Medium is Microsoft's latest SOTA model.",
     "format": "gguf",
     "settings": {
       "ctx_len": 128000,
-      "prompt_template": "<|user|>\n{prompt}<|end|>\n<|assistant|>\n",
+      "prompt_template": "<|user|> {prompt}<|end|><|assistant|><|end|>",
       "llama_model_path": "Phi-3-medium-128k-instruct-Q4_K_M.gguf",
       "ngl": 33
     },
diff --git a/extensions/inference-nitro-extension/resources/models/qwen2-7b/model.json b/extensions/inference-nitro-extension/resources/models/qwen2-7b/model.json
index 8939a98f3..a7613982c 100644
--- a/extensions/inference-nitro-extension/resources/models/qwen2-7b/model.json
+++ b/extensions/inference-nitro-extension/resources/models/qwen2-7b/model.json
@@ -7,8 +7,8 @@
     ],
     "id": "qwen2-7b",
     "object": "model",
-    "name": "Qwen 2 Instruct 7B Q4",
-    "version": "1.1",
+    "name": "Qwen 2 7B Instruct Q4",
+    "version": "1.2",
     "description": "Qwen is optimized at Chinese, ideal for everyday tasks.",
     "format": "gguf",
     "settings": {