diff --git a/web-app/src/containers/DropdownModelProvider.tsx b/web-app/src/containers/DropdownModelProvider.tsx
index 75a7d71d7..d34228daf 100644
--- a/web-app/src/containers/DropdownModelProvider.tsx
+++ b/web-app/src/containers/DropdownModelProvider.tsx
@@ -25,6 +25,7 @@ const DropdownModelProvider = ({ model }: DropdownModelProviderProps) => {
     providers,
     getProviderByName,
     selectModelProvider,
+    getModelBy,
     selectedProvider,
     selectedModel,
   } = useModelProvider()
@@ -52,8 +53,11 @@ const DropdownModelProvider = ({ model }: DropdownModelProviderProps) => {
     }
   }, [selectedProvider, selectedModel])
 
-  if (!providers.length) return null
+  const currentModel = selectedModel?.id
+    ? getModelBy(selectedModel?.id)
+    : undefined
 
+  if (!providers.length) return null
 
   const provider = getProviderByName(selectedProvider)
 
@@ -81,9 +85,9 @@ const DropdownModelProvider = ({ model }: DropdownModelProviderProps) => {
               </span>
             </button>
           </DropdownMenuTrigger>
-          {selectedModel && (
+          {currentModel && (
             <ModelSetting
-              model={selectedModel as Model}
+              model={currentModel as Model}
               provider={provider as ProviderObject}
             />
           )}
diff --git a/web-app/src/containers/ModelSetting.tsx b/web-app/src/containers/ModelSetting.tsx
index 7ccbc794b..e7a9d0d08 100644
--- a/web-app/src/containers/ModelSetting.tsx
+++ b/web-app/src/containers/ModelSetting.tsx
@@ -57,11 +57,19 @@ export function ModelSetting({ model, provider }: ModelSettingProps) {
         models: updatedModels,
       })
 
+      const params = Object.entries(updatedModel.settings).reduce(
+        (acc, [key, value]) => {
+          const rawVal = value.controller_props?.value
+          const num = parseFloat(rawVal as string)
+          acc[key] = !isNaN(num) ? num : rawVal
+          return acc
+        },
+        {} as Record<string, unknown>
+      ) as ModelSettingParams
       updateModel({
         id: model.id,
-        settings: Object.entries(updatedModel.settings).map(([key, value]) => ({
-          [key]: value.controller_props?.value,
-        })) as ModelSettingParams,
+        settings: params,
+        ...(params as unknown as object),
       })
     }
   }
@@ -75,7 +83,7 @@ export function ModelSetting({ model, provider }: ModelSettingProps) {
       </SheetTrigger>
       <SheetContent className="h-[calc(100%-8px)] top-1 right-1 rounded-e-md overflow-y-auto">
         <SheetHeader>
-          <SheetTitle>Model Setting {model.id}</SheetTitle>
+          <SheetTitle>Model Settings - {model.id}</SheetTitle>
           <SheetDescription>
             Configure model settings to optimize performance and behavior.
           </SheetDescription>
diff --git a/web-app/src/hooks/useChat.ts b/web-app/src/hooks/useChat.ts
index 2bff798a5..85e7a88f2 100644
--- a/web-app/src/hooks/useChat.ts
+++ b/web-app/src/hooks/useChat.ts
@@ -180,14 +180,14 @@ export const useChat = () => {
       resetTokenSpeed,
       provider,
       getMessages,
+      setAbortController,
       updateStreamingContent,
       addMessage,
       setPrompt,
       selectedModel,
-      currentAssistant?.instructions,
-      setAbortController,
-      updateLoadingModel,
+      currentAssistant,
       tools,
+      updateLoadingModel,
       updateTokenSpeed,
     ]
   )
diff --git a/web-app/src/hooks/useModelProvider.ts b/web-app/src/hooks/useModelProvider.ts
index 8fa4b976b..49bbbee39 100644
--- a/web-app/src/hooks/useModelProvider.ts
+++ b/web-app/src/hooks/useModelProvider.ts
@@ -6,6 +6,7 @@ type ModelProviderState = {
   providers: ModelProvider[]
   selectedProvider: string
   selectedModel: Model | null
+  getModelBy: (modelId: string) => Model | undefined
   setProviders: (providers: ModelProvider[]) => void
   getProviderByName: (providerName: string) => ModelProvider | undefined
   updateProvider: (providerName: string, data: Partial<ModelProvider>) => void
@@ -24,6 +25,13 @@ export const useModelProvider = create<ModelProviderState>()(
       providers: [],
       selectedProvider: 'llama.cpp',
       selectedModel: null,
+      getModelBy: (modelId: string) => {
+        const provider = get().providers.find(
+          (provider) => provider.provider === get().selectedProvider
+        )
+        if (!provider) return undefined
+        return provider.models.find((model) => model.id === modelId)
+      },
       setProviders: (providers) =>
         set((state) => {
           const existingProviders = state.providers
diff --git a/web-app/src/lib/predefined.ts b/web-app/src/lib/predefined.ts
index b3901f627..4fdc24cb9 100644
--- a/web-app/src/lib/predefined.ts
+++ b/web-app/src/lib/predefined.ts
@@ -1,6 +1,6 @@
 export const modelSettings = {
-  ctx_size: {
-    key: 'ctx_size',
+  ctx_len: {
+    key: 'ctx_len',
     title: 'Context Size',
     description: 'Size of the prompt context (0 = loaded from model).',
     controller_type: 'input',
@@ -10,290 +10,108 @@ export const modelSettings = {
       type: 'number',
     },
   },
-  n_predict: {
-    key: 'n_predict',
-    title: 'Max Tokens to Predict',
-    description: 'Maximum number of tokens to generate (-1 = infinity).',
-    controller_type: 'input',
-    controller_props: {
-      value: -1,
-      placeholder: '-1',
-      type: 'number',
-    },
-  },
-  batch_size: {
-    key: 'batch_size',
-    title: 'Batch Size',
-    description: 'Logical maximum batch size for processing prompts.',
-    controller_type: 'input',
-    controller_props: {
-      value: 2048,
-      placeholder: '2048',
-      type: 'number',
-    },
-  },
-  ubatch_size: {
-    key: 'ubatch_size',
-    title: 'uBatch Size',
-    description: 'Physical maximum batch size for processing prompts.',
-    controller_type: 'input',
-    controller_props: {
-      value: 512,
-      placeholder: '512',
-      type: 'number',
-    },
-  },
-  n_gpu_layers: {
-    key: 'n_gpu_layers',
+  ngl: {
+    key: 'ngl',
     title: 'GPU Layers',
     description:
       'Number of model layers to offload to the GPU (-1 for all layers, 0 for CPU only).',
     controller_type: 'input',
     controller_props: {
-      value: -1,
+      value: 100,
       placeholder: '-1',
       type: 'number',
     },
   },
-  device: {
-    key: 'device',
-    title: 'Devices for Offload',
-    description:
-      "Comma-separated list of devices to use for offloading (e.g., 'cuda:0', 'cuda:0,cuda:1'). Leave empty to use default/CPU only.",
+
+  temperature: {
+    key: 'temp',
+    title: 'Temperature',
+    description: 'Temperature for sampling (higher = more random).',
     controller_type: 'input',
     controller_props: {
-      value: '',
-      placeholder: 'cuda:0',
-      type: 'text',
+      value: 0.8,
+      placeholder: '0.8',
+      type: 'number',
+      min: 0,
+      step: 0.01,
     },
   },
-  split_mode: {
-    key: 'split_mode',
-    title: 'GPU Split Mode',
-    description: 'How to split the model across multiple GPUs.',
-    controller_type: 'dropdown',
-    controller_props: {
-      value: 'layer',
-      options: [
-        { value: 'none', name: 'None' },
-        { value: 'layer', name: 'Layer' },
-        { value: 'row', name: 'Row' },
-      ],
-    },
-  },
-  main_gpu: {
-    key: 'main_gpu',
-    title: 'Main GPU Index',
-    description:
-      'The GPU to use for the model (split-mode=none) or intermediate results (split-mode=row).',
+  top_k: {
+    key: 'top_k',
+    title: 'Top K',
+    description: 'Top-K sampling (0 = disabled).',
     controller_type: 'input',
     controller_props: {
-      value: 0,
-      placeholder: '0',
+      value: 40,
+      placeholder: '40',
       type: 'number',
     },
   },
-  flash_attn: {
-    key: 'flash_attn',
-    title: 'Flash Attention',
-    description: 'Enable Flash Attention for optimized performance.',
-    controller_type: 'checkbox',
+  top_p: {
+    key: 'top_p',
+    title: 'Top P',
+    description: 'Top-P sampling (1.0 = disabled).',
+    controller_type: 'input',
     controller_props: {
-      value: false,
+      value: 0.9,
+      placeholder: '0.9',
+      type: 'number',
     },
   },
-  cont_batching: {
-    key: 'cont_batching',
-    title: 'Continuous Batching',
-    description:
-      'Enable continuous batching (a.k.a dynamic batching) for concurrent requests (default: enabled).',
-    controller_type: 'checkbox',
-    controller_props: {
-      value: true,
-    },
-  },
-  no_mmap: {
-    key: 'no_mmap',
-    title: 'Disable mmap',
-    description:
-      'Do not memory-map model (slower load but may reduce pageouts if not using mlock).',
-    controller_type: 'checkbox',
-    controller_props: {
-      value: false,
-    },
-  },
-  mlock: {
-    key: 'mlock',
-    title: 'MLock',
-    description:
-      'Force system to keep model in RAM, preventing swapping/compression.',
-    controller_type: 'checkbox',
-    controller_props: {
-      value: false,
-    },
-  },
-  no_kv_offload: {
-    key: 'no_kv_offload',
-    title: 'Disable KV Offload',
-    description: 'Disable KV cache offload to GPU (if GPU is used).',
-    controller_type: 'checkbox',
-    controller_props: {
-      value: false,
-    },
-  },
-  cache_type_k: {
-    key: 'cache_type_k',
-    title: 'KV Cache K Type',
-    description: 'KV cache data type for Keys (default: f16).',
-    controller_type: 'dropdown',
-    controller_props: {
-      value: 'f16',
-      options: [
-        { value: 'f32', name: 'f32' },
-        { value: 'f16', name: 'f16' },
-        { value: 'bf16', name: 'bf16' },
-        { value: 'q8_0', name: 'q8_0' },
-        { value: 'q4_0', name: 'q4_0' },
-        { value: 'q4_1', name: 'q4_1' },
-        { value: 'iq4_nl', name: 'iq4_nl' },
-        { value: 'q5_0', name: 'q5_0' },
-        { value: 'q5_1', name: 'q5_1' },
-      ],
-    },
-  },
-  cache_type_v: {
-    key: 'cache_type_v',
-    title: 'KV Cache V Type',
-    description: 'KV cache data type for Values (default: f16).',
-    controller_type: 'dropdown',
-    controller_props: {
-      value: 'f16',
-      options: [
-        { value: 'f32', name: 'f32' },
-        { value: 'f16', name: 'f16' },
-        { value: 'bf16', name: 'bf16' },
-        { value: 'q8_0', name: 'q8_0' },
-        { value: 'q4_0', name: 'q4_0' },
-        { value: 'q4_1', name: 'q4_1' },
-        { value: 'iq4_nl', name: 'iq4_nl' },
-        { value: 'q5_0', name: 'q5_0' },
-        { value: 'q5_1', name: 'q5_1' },
-      ],
-    },
-  },
-  defrag_thold: {
-    key: 'defrag_thold',
-    title: 'KV Cache Defragmentation Threshold',
-    description: 'Threshold for KV cache defragmentation (< 0 to disable).',
+  min_p: {
+    key: 'min_p',
+    title: 'Min P',
+    description: 'Min-P sampling (0.0 = disabled).',
     controller_type: 'input',
     controller_props: {
       value: 0.1,
       placeholder: '0.1',
       type: 'number',
-      step: 0.01,
     },
   },
-  rope_scaling: {
-    key: 'rope_scaling',
-    title: 'RoPE Scaling Method',
-    description: 'RoPE frequency scaling method.',
-    controller_type: 'dropdown',
-    controller_props: {
-      value: 'none',
-      options: [
-        { value: 'none', name: 'None' },
-        { value: 'linear', name: 'Linear' },
-        { value: 'yarn', name: 'YaRN' },
-      ],
-    },
-  },
-  rope_scale: {
-    key: 'rope_scale',
-    title: 'RoPE Scale Factor',
-    description: 'RoPE context scaling factor.',
-    controller_type: 'input',
-    controller_props: {
-      value: 1.0,
-      placeholder: '1.0',
-      type: 'number',
-      min: 0,
-      step: 0.01,
-    },
-  },
-  rope_freq_base: {
-    key: 'rope_freq_base',
-    title: 'RoPE Frequency Base',
-    description: 'RoPE base frequency (0 = loaded from model).',
-    controller_type: 'input',
-    controller_props: {
-      value: 0,
-      placeholder: '0 (model default)',
-      type: 'number',
-    },
-  },
-  rope_freq_scale: {
-    key: 'rope_freq_scale',
-    title: 'RoPE Frequency Scale Factor',
-    description: 'RoPE frequency scaling factor.',
-    controller_type: 'input',
-    controller_props: {
-      value: 1.0,
-      placeholder: '1.0',
-      type: 'number',
-      min: 0,
-      step: 0.01,
-    },
-  },
-  mirostat: {
-    key: 'mirostat',
-    title: 'Mirostat Mode',
+  repeat_last_n: {
+    key: 'repeat_last_n',
+    title: 'Repeat Last N',
     description:
-      'Use Mirostat sampling (0: disabled, 1: Mirostat V1, 2: Mirostat V2).',
-    controller_type: 'dropdown',
-    controller_props: {
-      value: 0,
-      options: [
-        { value: 0, name: 'Disabled' },
-        { value: 1, name: 'Mirostat V1' },
-        { value: 2, name: 'Mirostat V2' },
-      ],
-    },
-  },
-  mirostat_lr: {
-    key: 'mirostat_lr',
-    title: 'Mirostat Learning Rate',
-    description: 'Mirostat learning rate (eta).',
+      'Number of tokens to consider for repeat penalty (0 = disabled, -1 = ctx_size).',
     controller_type: 'input',
     controller_props: {
-      value: 0.1,
-      placeholder: '0.1',
+      value: 64,
+      placeholder: '64',
       type: 'number',
-      min: 0,
-      step: 0.01,
     },
   },
-  mirostat_ent: {
-    key: 'mirostat_ent',
-    title: 'Mirostat Target Entropy',
-    description: 'Mirostat target entropy (tau).',
+  repeat_penalty: {
+    key: 'repeat_penalty',
+    title: 'Repeat Penalty',
+    description: 'Penalize repeating token sequences (1.0 = disabled).',
     controller_type: 'input',
     controller_props: {
-      value: 5.0,
-      placeholder: '5.0',
+      value: 1.0,
+      placeholder: '1.0',
       type: 'number',
-      min: 0,
-      step: 0.01,
     },
   },
-  json_schema_file: {
-    key: 'json_schema_file',
-    title: 'JSON Schema File',
-    description: 'Path to a JSON schema file to constrain generations.',
+  presence_penalty: {
+    key: 'presence_penalty',
+    title: 'Presence Penalty',
+    description: 'Repeat alpha presence penalty (0.0 = disabled).',
     controller_type: 'input',
     controller_props: {
-      value: '',
-      placeholder: 'path/to/schema.json',
-      type: 'text',
+      value: 0.0,
+      placeholder: '0.0',
+      type: 'number',
+    },
+  },
+  frequency_penalty: {
+    key: 'frequency_penalty',
+    title: 'Frequency Penalty',
+    description: 'Repeat alpha frequency penalty (0.0 = disabled).',
+    controller_type: 'input',
+    controller_props: {
+      value: 0.0,
+      placeholder: '0.0',
+      type: 'number',
     },
   },
 }