Merge pull request #23 from lordmathis/feat/start-on-request

feat: On-Demand Instance Start
Remove unused import from zodFormUtils.ts
2025-11-06 00:54:23 +00:00 · 2025-08-20 16:04:59 +02:00 · 2025-08-20 16:03:09 +02:00 · 2025-08-20 15:58:08 +02:00 · 2025-08-20 14:56:11 +02:00 · 2025-08-20 14:41:42 +02:00
31 changed files with 1297 additions and 933 deletions
--- a/.github/workflows/release.yaml
+++ b/.github/workflows/release.yaml
@@ -29,6 +29,8 @@ jobs:
          npm ci

      - name: Build Web UI
+        env:
+          VITE_APP_VERSION: ${{ github.ref_name }}
        run: |
          cd webui
          npm run build
--- a/1
+++ b/1
@@ -0,0 +1 @@
+llamactl.org
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -0,0 +1,138 @@
+# Contributing to Llamactl
+
+Thank you for considering contributing to Llamactl! This document outlines the development setup and contribution process.
+
+## Development Setup
+
+### Prerequisites
+
+- Go 1.24 or later
+- Node.js 22 or later
+- `llama-server` executable (from [llama.cpp](https://github.com/ggml-org/llama.cpp))
+
+### Getting Started
+
+1. **Clone the repository**
+   ```bash
+   git clone https://github.com/lordmathis/llamactl.git
+   cd llamactl
+   ```
+
+2. **Install dependencies**
+   ```bash
+   # Go dependencies
+   go mod download
+   
+   # Frontend dependencies
+   cd webui && npm ci && cd ..
+   ```
+
+3. **Run for development**
+   ```bash
+   # Start backend server
+   go run ./cmd/server
+   ```
+   Server will be available at `http://localhost:8080`
+   
+   ```bash
+   # In a separate terminal, start frontend dev server
+   cd webui && npm run dev
+   ```
+   Development UI will be available at `http://localhost:5173`
+
+4. **Common development commands**
+   ```bash
+   # Backend
+   go test ./... -v                    # Run tests
+   go test -race ./... -v              # Run with race detector
+   go fmt ./... && go vet ./...        # Format and vet code
+   
+   # Frontend (run from webui/ directory)
+   npm run test:run                    # Run tests once
+   npm run test                        # Run tests in watch mode
+   npm run type-check                  # TypeScript type checking
+   npm run lint:fix                    # Lint and fix issues
+   ```
+
+## Before Submitting a Pull Request
+
+### Required Checks
+
+All the following must pass:
+
+1. **Backend**
+   ```bash
+   go test ./... -v
+   go test -race ./... -v
+   go fmt ./... && go vet ./...
+   go build -o llamactl ./cmd/server
+   ```
+
+2. **Frontend**
+   ```bash
+   cd webui
+   npm run test:run
+   npm run type-check
+   npm run build
+   ```
+
+### API Documentation
+
+If changes affect API endpoints, update Swagger documentation:
+
+```bash
+# Install swag if needed
+go install github.com/swaggo/swag/cmd/swag@latest
+
+# Update Swagger comments in pkg/server/handlers.go
+# Then regenerate docs
+swag init -g cmd/server/main.go -o apidocs
+```
+
+## Pull Request Guidelines
+
+### Pull Request Titles
+Use this format for pull request titles:
+- `feat:` for new features
+- `fix:` for bug fixes  
+- `docs:` for documentation changes
+- `test:` for test additions or modifications
+- `refactor:` for code refactoring
+
+### Submission Process
+1. Create a feature branch from `main`
+2. Make changes following the coding standards
+3. Run all required checks listed above
+4. Update documentation if necessary
+5. Submit pull request with:
+   - Clear description of changes
+   - Reference to any related issues
+   - Screenshots for UI changes
+
+## Code Style and Testing
+
+### Testing Strategy
+- Backend tests use Go's built-in testing framework
+- Frontend tests use Vitest and React Testing Library
+- Run tests frequently during development
+- Add tests for new features and bug fixes
+
+### Go
+- Follow standard Go formatting (`go fmt`)
+- Use meaningful variable and function names  
+- Add comments for exported functions and types
+- Handle errors appropriately
+
+### TypeScript/React
+- Use TypeScript strictly (avoid `any` when possible)
+- Follow React hooks best practices
+- Use meaningful component and variable names
+- Prefer functional components over class components
+
+## Getting Help
+
+- Check existing [issues](https://github.com/lordmathis/llamactl/issues)
+- Review the [README.md](README.md) for usage documentation  
+- Look at existing code for patterns and conventions
+
+Thank you for contributing to Llamactl!
--- a/README.md
+++ b/README.md
@@ -2,7 +2,7 @@

 ![Build and Release](https://github.com/lordmathis/llamactl/actions/workflows/release.yaml/badge.svg) ![Go Tests](https://github.com/lordmathis/llamactl/actions/workflows/go_test.yaml/badge.svg) ![WebUI Tests](https://github.com/lordmathis/llamactl/actions/workflows/webui_test.yaml/badge.svg)

-**Management server for multiple llama.cpp instances with OpenAI-compatible API routing.**
+**Management server and proxy for multiple llama.cpp instances with OpenAI-compatible API routing.**

 ## Why llamactl?

@@ -11,7 +11,11 @@
 🌐 **Web Dashboard**: Modern React UI for visual management (unlike CLI-only tools)  
 🔐 **API Key Authentication**: Separate keys for management vs inference access  
 📊 **Instance Monitoring**: Health checks, auto-restart, log management  
-⚡ **Persistent State**: Instances survive server restarts
+⏳ **Idle Timeout Management**: Automatically stop idle instances after a configurable period  
+💡 **On-Demand Instance Start**: Automatically launch instances upon receiving OpenAI-compatible API requests  
+💾 **State Persistence**: Ensure instances remain intact across server restarts  
+
+![Dashboard Screenshot](docs/images/screenshot.png)

 **Choose llamactl if**: You need authentication, health monitoring, auto-restart, and centralized management of multiple llama-server instances  
 **Choose Ollama if**: You want the simplest setup with strong community ecosystem and third-party integrations  
@@ -113,6 +117,10 @@ instances:
  default_auto_restart: true     # Auto-restart new instances by default
  default_max_restarts: 3        # Max restarts for new instances
  default_restart_delay: 5       # Restart delay (seconds) for new instances
+  default_on_demand_start: true  # Default on-demand start setting
+  on_demand_start_timeout: 120   # Default on-demand start timeout in seconds
+  timeout_check_interval: 5      # Idle instance timeout check in minutes
+

 auth:
  require_inference_auth: true   # Require auth for inference endpoints
@@ -170,16 +178,19 @@ server:

 ```yaml
 instances:
-  port_range: [8000, 9000]           # Port range for instances (default: [8000, 9000])
-  data_dir: "~/.local/share/llamactl" # Directory for all llamactl data (default varies by OS)
-  configs_dir: "~/.local/share/llamactl/instances" # Directory for instance configs (default: data_dir/instances)
-  logs_dir: "~/.local/share/llamactl/logs" # Directory for instance logs (default: data_dir/logs)
-  auto_create_dirs: true             # Automatically create data/config/logs directories (default: true)
-  max_instances: -1                  # Maximum instances (-1 = unlimited)
-  llama_executable: "llama-server"   # Path to llama-server executable
-  default_auto_restart: true         # Default auto-restart setting
-  default_max_restarts: 3            # Default maximum restart attempts
-  default_restart_delay: 5           # Default restart delay in seconds
+  port_range: [8000, 9000]                          # Port range for instances (default: [8000, 9000])
+  data_dir: "~/.local/share/llamactl"               # Directory for all llamactl data (default varies by OS)
+  configs_dir: "~/.local/share/llamactl/instances"  # Directory for instance configs (default: data_dir/instances)
+  logs_dir: "~/.local/share/llamactl/logs"          # Directory for instance logs (default: data_dir/logs)
+  auto_create_dirs: true                            # Automatically create data/config/logs directories (default: true)
+  max_instances: -1                                 # Maximum instances (-1 = unlimited)
+  llama_executable: "llama-server"                  # Path to llama-server executable
+  default_auto_restart: true                        # Default auto-restart setting
+  default_max_restarts: 3                           # Default maximum restart attempts
+  default_restart_delay: 5                          # Default restart delay in seconds
+  default_on_demand_start: true                     # Default on-demand start setting
+  on_demand_start_timeout: 120                      # Default on-demand start timeout in seconds
+  timeout_check_interval: 5                         # Default instance timeout check interval in minutes
 ```

 **Environment Variables:**
@@ -193,6 +204,10 @@ instances:
 - `LLAMACTL_DEFAULT_AUTO_RESTART` - Default auto-restart setting (true/false)
 - `LLAMACTL_DEFAULT_MAX_RESTARTS` - Default maximum restarts
 - `LLAMACTL_DEFAULT_RESTART_DELAY` - Default restart delay in seconds
+- `LLAMACTL_DEFAULT_ON_DEMAND_START` - Default on-demand start setting (true/false)
+- `LLAMACTL_ON_DEMAND_START_TIMEOUT` - Default on-demand start timeout in seconds
+- `LLAMACTL_TIMEOUT_CHECK_INTERVAL` - Default instance timeout check interval in minutes
+

 #### Authentication Configuration

--- a/apidocs/docs.go
+++ b/apidocs/docs.go
@@ -1,5 +1,5 @@
-// Package docs Code generated by swaggo/swag. DO NOT EDIT
-package docs
+// Package apidocs Code generated by swaggo/swag. DO NOT EDIT
+package apidocs

 import "github.com/swaggo/swag"

@@ -37,7 +37,7 @@ const docTemplate = `{
                        "schema": {
                            "type": "array",
                            "items": {
-                                "$ref": "#/definitions/llamactl.Instance"
+                                "$ref": "#/definitions/instance.Process"
                            }
                        }
                    },
@@ -75,7 +75,7 @@ const docTemplate = `{
                    "200": {
                        "description": "Instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -120,7 +120,7 @@ const docTemplate = `{
                        "in": "body",
                        "required": true,
                        "schema": {
-                            "$ref": "#/definitions/llamactl.CreateInstanceOptions"
+                            "$ref": "#/definitions/instance.CreateInstanceOptions"
                        }
                    }
                ],
@@ -128,7 +128,7 @@ const docTemplate = `{
                    "200": {
                        "description": "Updated instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -173,7 +173,7 @@ const docTemplate = `{
                        "in": "body",
                        "required": true,
                        "schema": {
-                            "$ref": "#/definitions/llamactl.CreateInstanceOptions"
+                            "$ref": "#/definitions/instance.CreateInstanceOptions"
                        }
                    }
                ],
@@ -181,7 +181,7 @@ const docTemplate = `{
                    "201": {
                        "description": "Created instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -401,7 +401,7 @@ const docTemplate = `{
                    "200": {
                        "description": "Restarted instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -444,7 +444,7 @@ const docTemplate = `{
                    "200": {
                        "description": "Started instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -487,7 +487,7 @@ const docTemplate = `{
                    "200": {
                        "description": "Stopped instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -639,7 +639,35 @@ const docTemplate = `{
                    "200": {
                        "description": "List of OpenAI-compatible instances",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.OpenAIListInstancesResponse"
+                            "$ref": "#/definitions/server.OpenAIListInstancesResponse"
+                        }
+                    },
+                    "500": {
+                        "description": "Internal Server Error",
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                }
+            }
+        },
+        "/version": {
+            "get": {
+                "security": [
+                    {
+                        "ApiKeyAuth": []
+                    }
+                ],
+                "description": "Returns the version of the llamactl command",
+                "tags": [
+                    "version"
+                ],
+                "summary": "Get llamactl version",
+                "responses": {
+                    "200": {
+                        "description": "Version information",
+                        "schema": {
+                            "type": "string"
                        }
                    },
                    "500": {
@@ -653,7 +681,7 @@ const docTemplate = `{
        }
    },
    "definitions": {
-        "llamactl.CreateInstanceOptions": {
+        "instance.CreateInstanceOptions": {
            "type": "object",
            "properties": {
                "alias": {
@@ -751,7 +779,6 @@ const docTemplate = `{
                    "type": "string"
                },
                "draft_max": {
-                    "description": "Speculative decoding params",
                    "type": "integer"
                },
                "draft_min": {
@@ -955,7 +982,7 @@ const docTemplate = `{
                    "type": "boolean"
                },
                "no_context_shift": {
-                    "description": "Server/Example-specific params",
+                    "description": "Example-specific params",
                    "type": "boolean"
                },
                "no_escape": {
@@ -1027,10 +1054,10 @@ const docTemplate = `{
                "presence_penalty": {
                    "type": "number"
                },
-                "priority": {
+                "prio": {
                    "type": "integer"
                },
-                "priority_batch": {
+                "prio_batch": {
                    "type": "integer"
                },
                "props": {
@@ -1101,7 +1128,7 @@ const docTemplate = `{
                "ssl_key_file": {
                    "type": "string"
                },
-                "temperature": {
+                "temp": {
                    "type": "number"
                },
                "tensor_split": {
@@ -1167,7 +1194,7 @@ const docTemplate = `{
                }
            }
        },
-        "llamactl.Instance": {
+        "instance.Process": {
            "type": "object",
            "properties": {
                "created": {
@@ -1183,7 +1210,7 @@ const docTemplate = `{
                }
            }
        },
-        "llamactl.OpenAIInstance": {
+        "server.OpenAIInstance": {
            "type": "object",
            "properties": {
                "created": {
@@ -1200,13 +1227,13 @@ const docTemplate = `{
                }
            }
        },
-        "llamactl.OpenAIListInstancesResponse": {
+        "server.OpenAIListInstancesResponse": {
            "type": "object",
            "properties": {
                "data": {
                    "type": "array",
                    "items": {
-                        "$ref": "#/definitions/llamactl.OpenAIInstance"
+                        "$ref": "#/definitions/server.OpenAIInstance"
                    }
                },
                "object": {
--- a/apidocs/swagger.json
+++ b/apidocs/swagger.json
@@ -30,7 +30,7 @@
                        "schema": {
                            "type": "array",
                            "items": {
-                                "$ref": "#/definitions/llamactl.Instance"
+                                "$ref": "#/definitions/instance.Process"
                            }
                        }
                    },
@@ -68,7 +68,7 @@
                    "200": {
                        "description": "Instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -113,7 +113,7 @@
                        "in": "body",
                        "required": true,
                        "schema": {
-                            "$ref": "#/definitions/llamactl.CreateInstanceOptions"
+                            "$ref": "#/definitions/instance.CreateInstanceOptions"
                        }
                    }
                ],
@@ -121,7 +121,7 @@
                    "200": {
                        "description": "Updated instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -166,7 +166,7 @@
                        "in": "body",
                        "required": true,
                        "schema": {
-                            "$ref": "#/definitions/llamactl.CreateInstanceOptions"
+                            "$ref": "#/definitions/instance.CreateInstanceOptions"
                        }
                    }
                ],
@@ -174,7 +174,7 @@
                    "201": {
                        "description": "Created instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -394,7 +394,7 @@
                    "200": {
                        "description": "Restarted instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -437,7 +437,7 @@
                    "200": {
                        "description": "Started instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -480,7 +480,7 @@
                    "200": {
                        "description": "Stopped instance details",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.Instance"
+                            "$ref": "#/definitions/instance.Process"
                        }
                    },
                    "400": {
@@ -632,7 +632,35 @@
                    "200": {
                        "description": "List of OpenAI-compatible instances",
                        "schema": {
-                            "$ref": "#/definitions/llamactl.OpenAIListInstancesResponse"
+                            "$ref": "#/definitions/server.OpenAIListInstancesResponse"
+                        }
+                    },
+                    "500": {
+                        "description": "Internal Server Error",
+                        "schema": {
+                            "type": "string"
+                        }
+                    }
+                }
+            }
+        },
+        "/version": {
+            "get": {
+                "security": [
+                    {
+                        "ApiKeyAuth": []
+                    }
+                ],
+                "description": "Returns the version of the llamactl command",
+                "tags": [
+                    "version"
+                ],
+                "summary": "Get llamactl version",
+                "responses": {
+                    "200": {
+                        "description": "Version information",
+                        "schema": {
+                            "type": "string"
                        }
                    },
                    "500": {
@@ -646,7 +674,7 @@
        }
    },
    "definitions": {
-        "llamactl.CreateInstanceOptions": {
+        "instance.CreateInstanceOptions": {
            "type": "object",
            "properties": {
                "alias": {
@@ -744,7 +772,6 @@
                    "type": "string"
                },
                "draft_max": {
-                    "description": "Speculative decoding params",
                    "type": "integer"
                },
                "draft_min": {
@@ -948,7 +975,7 @@
                    "type": "boolean"
                },
                "no_context_shift": {
-                    "description": "Server/Example-specific params",
+                    "description": "Example-specific params",
                    "type": "boolean"
                },
                "no_escape": {
@@ -1020,10 +1047,10 @@
                "presence_penalty": {
                    "type": "number"
                },
-                "priority": {
+                "prio": {
                    "type": "integer"
                },
-                "priority_batch": {
+                "prio_batch": {
                    "type": "integer"
                },
                "props": {
@@ -1094,7 +1121,7 @@
                "ssl_key_file": {
                    "type": "string"
                },
-                "temperature": {
+                "temp": {
                    "type": "number"
                },
                "tensor_split": {
@@ -1160,7 +1187,7 @@
                }
            }
        },
-        "llamactl.Instance": {
+        "instance.Process": {
            "type": "object",
            "properties": {
                "created": {
@@ -1176,7 +1203,7 @@
                }
            }
        },
-        "llamactl.OpenAIInstance": {
+        "server.OpenAIInstance": {
            "type": "object",
            "properties": {
                "created": {
@@ -1193,13 +1220,13 @@
                }
            }
        },
-        "llamactl.OpenAIListInstancesResponse": {
+        "server.OpenAIListInstancesResponse": {
            "type": "object",
            "properties": {
                "data": {
                    "type": "array",
                    "items": {
-                        "$ref": "#/definitions/llamactl.OpenAIInstance"
+                        "$ref": "#/definitions/server.OpenAIInstance"
                    }
                },
                "object": {
--- a/apidocs/swagger.yaml
+++ b/apidocs/swagger.yaml
@@ -1,6 +1,6 @@
 basePath: /api/v1
 definitions:
-  llamactl.CreateInstanceOptions:
+  instance.CreateInstanceOptions:
    properties:
      alias:
        type: string
@@ -66,7 +66,6 @@ definitions:
      device_draft:
        type: string
      draft_max:
-        description: Speculative decoding params
        type: integer
      draft_min:
        type: integer
@@ -203,7 +202,7 @@ definitions:
      no_cont_batching:
        type: boolean
      no_context_shift:
-        description: Server/Example-specific params
+        description: Example-specific params
        type: boolean
      no_escape:
        type: boolean
@@ -251,9 +250,9 @@ definitions:
        type: integer
      presence_penalty:
        type: number
-      priority:
+      prio:
        type: integer
-      priority_batch:
+      prio_batch:
        type: integer
      props:
        type: boolean
@@ -301,7 +300,7 @@ definitions:
        type: string
      ssl_key_file:
        type: string
-      temperature:
+      temp:
        type: number
      tensor_split:
        type: string
@@ -345,7 +344,7 @@ definitions:
      yarn_orig_ctx:
        type: integer
    type: object
-  llamactl.Instance:
+  instance.Process:
    properties:
      created:
        description: Creation time
@@ -356,7 +355,7 @@ definitions:
        description: Status
        type: boolean
    type: object
-  llamactl.OpenAIInstance:
+  server.OpenAIInstance:
    properties:
      created:
        type: integer
@@ -367,11 +366,11 @@ definitions:
      owned_by:
        type: string
    type: object
-  llamactl.OpenAIListInstancesResponse:
+  server.OpenAIListInstancesResponse:
    properties:
      data:
        items:
-          $ref: '#/definitions/llamactl.OpenAIInstance'
+          $ref: '#/definitions/server.OpenAIInstance'
        type: array
      object:
        type: string
@@ -393,7 +392,7 @@ paths:
          description: List of instances
          schema:
            items:
-              $ref: '#/definitions/llamactl.Instance'
+              $ref: '#/definitions/instance.Process'
            type: array
        "500":
          description: Internal Server Error
@@ -441,7 +440,7 @@ paths:
        "200":
          description: Instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid name format
          schema:
@@ -470,12 +469,12 @@ paths:
        name: options
        required: true
        schema:
-          $ref: '#/definitions/llamactl.CreateInstanceOptions'
+          $ref: '#/definitions/instance.CreateInstanceOptions'
      responses:
        "201":
          description: Created instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid request body
          schema:
@@ -504,12 +503,12 @@ paths:
        name: options
        required: true
        schema:
-          $ref: '#/definitions/llamactl.CreateInstanceOptions'
+          $ref: '#/definitions/instance.CreateInstanceOptions'
      responses:
        "200":
          description: Updated instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid name format
          schema:
@@ -627,7 +626,7 @@ paths:
        "200":
          description: Restarted instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid name format
          schema:
@@ -654,7 +653,7 @@ paths:
        "200":
          description: Started instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid name format
          schema:
@@ -681,7 +680,7 @@ paths:
        "200":
          description: Stopped instance details
          schema:
-            $ref: '#/definitions/llamactl.Instance'
+            $ref: '#/definitions/instance.Process'
        "400":
          description: Invalid name format
          schema:
@@ -777,7 +776,7 @@ paths:
        "200":
          description: List of OpenAI-compatible instances
          schema:
-            $ref: '#/definitions/llamactl.OpenAIListInstancesResponse'
+            $ref: '#/definitions/server.OpenAIListInstancesResponse'
        "500":
          description: Internal Server Error
          schema:
@@ -787,4 +786,21 @@ paths:
      summary: List instances in OpenAI-compatible format
      tags:
      - openai
+  /version:
+    get:
+      description: Returns the version of the llamactl command
+      responses:
+        "200":
+          description: Version information
+          schema:
+            type: string
+        "500":
+          description: Internal Server Error
+          schema:
+            type: string
+      security:
+      - ApiKeyAuth: []
+      summary: Get llamactl version
+      tags:
+      - version
 swagger: "2.0"
--- a/cmd/server/main.go
+++ b/cmd/server/main.go
@@ -11,6 +11,11 @@ import (
 	"syscall"
 )

+// version is set at build time using -ldflags "-X main.version=1.0.0"
+var version string = "unknown"
+var commitHash string = "unknown"
+var buildTime string = "unknown"
+
 // @title llamactl API
 // @version 1.0
 // @description llamactl is a control server for managing Llama Server instances.
@@ -19,6 +24,14 @@ import (
 // @basePath /api/v1
 func main() {

+	// --version flag to print the version
+	if len(os.Args) > 1 && os.Args[1] == "--version" {
+		fmt.Printf("llamactl version: %s\n", version)
+		fmt.Printf("Commit hash: %s\n", commitHash)
+		fmt.Printf("Build time: %s\n", buildTime)
+		return
+	}
+
 	configPath := os.Getenv("LLAMACTL_CONFIG_PATH")
 	cfg, err := config.LoadConfig(configPath)
 	if err != nil {
@@ -26,6 +39,11 @@ func main() {
 		fmt.Println("Using default configuration.")
 	}

+	// Set version information
+	cfg.Version = version
+	cfg.CommitHash = commitHash
+	cfg.BuildTime = buildTime
+
 	// Create the data directory if it doesn't exist
 	if cfg.Instances.AutoCreateDirs {
 		if err := os.MkdirAll(cfg.Instances.InstancesDir, 0755); err != nil {
--- a/docs/images/screenshot.png
+++ b/docs/images/screenshot.png
--- a/pkg/config/config.go
+++ b/pkg/config/config.go
@@ -12,9 +12,12 @@ import (

 // AppConfig represents the configuration for llamactl
 type AppConfig struct {
-	Server    ServerConfig    `yaml:"server"`
-	Instances InstancesConfig `yaml:"instances"`
-	Auth      AuthConfig      `yaml:"auth"`
+	Server     ServerConfig    `yaml:"server"`
+	Instances  InstancesConfig `yaml:"instances"`
+	Auth       AuthConfig      `yaml:"auth"`
+	Version    string          `yaml:"-"`
+	CommitHash string          `yaml:"-"`
+	BuildTime  string          `yaml:"-"`
 }

 // ServerConfig contains HTTP server configuration
@@ -63,6 +66,15 @@ type InstancesConfig struct {

 	// Default restart delay for new instances (in seconds)
 	DefaultRestartDelay int `yaml:"default_restart_delay"`
+
+	// Default on-demand start setting for new instances
+	DefaultOnDemandStart bool `yaml:"default_on_demand_start"`
+
+	// How long to wait for an instance to start on demand (in seconds)
+	OnDemandStartTimeout int `yaml:"on_demand_start_timeout,omitempty"`
+
+	// Interval for checking instance timeouts (in minutes)
+	TimeoutCheckInterval int `yaml:"timeout_check_interval"`
 }

 // AuthConfig contains authentication settings
@@ -95,16 +107,19 @@ func LoadConfig(configPath string) (AppConfig, error) {
 			EnableSwagger:  false,
 		},
 		Instances: InstancesConfig{
-			PortRange:           [2]int{8000, 9000},
-			DataDir:             getDefaultDataDirectory(),
-			InstancesDir:        filepath.Join(getDefaultDataDirectory(), "instances"),
-			LogsDir:             filepath.Join(getDefaultDataDirectory(), "logs"),
-			AutoCreateDirs:      true,
-			MaxInstances:        -1, // -1 means unlimited
-			LlamaExecutable:     "llama-server",
-			DefaultAutoRestart:  true,
-			DefaultMaxRestarts:  3,
-			DefaultRestartDelay: 5,
+			PortRange:            [2]int{8000, 9000},
+			DataDir:              getDefaultDataDirectory(),
+			InstancesDir:         filepath.Join(getDefaultDataDirectory(), "instances"),
+			LogsDir:              filepath.Join(getDefaultDataDirectory(), "logs"),
+			AutoCreateDirs:       true,
+			MaxInstances:         -1, // -1 means unlimited
+			LlamaExecutable:      "llama-server",
+			DefaultAutoRestart:   true,
+			DefaultMaxRestarts:   3,
+			DefaultRestartDelay:  5,
+			DefaultOnDemandStart: true,
+			OnDemandStartTimeout: 120, // 2 minutes
+			TimeoutCheckInterval: 5,   // Check timeouts every 5 minutes
 		},
 		Auth: AuthConfig{
 			RequireInferenceAuth:  true,
@@ -214,6 +229,21 @@ func loadEnvVars(cfg *AppConfig) {
 			cfg.Instances.DefaultRestartDelay = seconds
 		}
 	}
+	if onDemandStart := os.Getenv("LLAMACTL_DEFAULT_ON_DEMAND_START"); onDemandStart != "" {
+		if b, err := strconv.ParseBool(onDemandStart); err == nil {
+			cfg.Instances.DefaultOnDemandStart = b
+		}
+	}
+	if onDemandTimeout := os.Getenv("LLAMACTL_ON_DEMAND_START_TIMEOUT"); onDemandTimeout != "" {
+		if seconds, err := strconv.Atoi(onDemandTimeout); err == nil {
+			cfg.Instances.OnDemandStartTimeout = seconds
+		}
+	}
+	if timeoutCheckInterval := os.Getenv("LLAMACTL_TIMEOUT_CHECK_INTERVAL"); timeoutCheckInterval != "" {
+		if minutes, err := strconv.Atoi(timeoutCheckInterval); err == nil {
+			cfg.Instances.TimeoutCheckInterval = minutes
+		}
+	}
 	// Auth config
 	if requireInferenceAuth := os.Getenv("LLAMACTL_REQUIRE_INFERENCE_AUTH"); requireInferenceAuth != "" {
 		if b, err := strconv.ParseBool(requireInferenceAuth); err == nil {
--- a/pkg/instance/instance.go
+++ b/pkg/instance/instance.go
@@ -13,16 +13,32 @@ import (
 	"net/url"
 	"os/exec"
 	"sync"
+	"sync/atomic"
 	"time"
 )

+// TimeProvider interface allows for testing with mock time
+type TimeProvider interface {
+	Now() time.Time
+}
+
+// realTimeProvider implements TimeProvider using the actual time
+type realTimeProvider struct{}
+
+func (realTimeProvider) Now() time.Time {
+	return time.Now()
+}
+
 type CreateInstanceOptions struct {
 	// Auto restart
-	AutoRestart *bool `json:"auto_restart,omitempty"`
-	MaxRestarts *int  `json:"max_restarts,omitempty"`
-	// RestartDelay duration in seconds
-	RestartDelay *int `json:"restart_delay_seconds,omitempty"`
-
+	AutoRestart  *bool `json:"auto_restart,omitempty"`
+	MaxRestarts  *int  `json:"max_restarts,omitempty"`
+	RestartDelay *int  `json:"restart_delay,omitempty"`
+	// On demand start
+	OnDemandStart *bool `json:"on_demand_start,omitempty"`
+	// Idle timeout
+	IdleTimeout *int `json:"idle_timeout,omitempty"`
+	// LlamaServerOptions contains the options for the llama server
 	llamacpp.LlamaServerOptions `json:",inline"`
 }

@@ -32,9 +48,11 @@ type CreateInstanceOptions struct {
 func (c *CreateInstanceOptions) UnmarshalJSON(data []byte) error {
 	// First, unmarshal into a temporary struct without the embedded type
 	type tempCreateOptions struct {
-		AutoRestart  *bool `json:"auto_restart,omitempty"`
-		MaxRestarts  *int  `json:"max_restarts,omitempty"`
-		RestartDelay *int  `json:"restart_delay_seconds,omitempty"`
+		AutoRestart   *bool `json:"auto_restart,omitempty"`
+		MaxRestarts   *int  `json:"max_restarts,omitempty"`
+		RestartDelay  *int  `json:"restart_delay,omitempty"`
+		OnDemandStart *bool `json:"on_demand_start,omitempty"`
+		IdleTimeout   *int  `json:"idle_timeout,omitempty"`
 	}

 	var temp tempCreateOptions
@@ -46,6 +64,8 @@ func (c *CreateInstanceOptions) UnmarshalJSON(data []byte) error {
 	c.AutoRestart = temp.AutoRestart
 	c.MaxRestarts = temp.MaxRestarts
 	c.RestartDelay = temp.RestartDelay
+	c.OnDemandStart = temp.OnDemandStart
+	c.IdleTimeout = temp.IdleTimeout

 	// Now unmarshal the embedded LlamaServerOptions
 	if err := json.Unmarshal(data, &c.LlamaServerOptions); err != nil {
@@ -83,6 +103,10 @@ type Process struct {
 	// Restart control
 	restartCancel context.CancelFunc `json:"-"` // Cancel function for pending restarts
 	monitorDone   chan struct{}      `json:"-"` // Channel to signal monitor goroutine completion
+
+	// Timeout management
+	lastRequestTime atomic.Int64 // Unix timestamp of last request
+	timeProvider    TimeProvider `json:"-"` // Time provider for testing
 }

 // validateAndCopyOptions validates and creates a deep copy of the provided options
@@ -117,6 +141,20 @@ func validateAndCopyOptions(name string, options *CreateInstanceOptions) *Create
 			}
 			optionsCopy.RestartDelay = &restartDelay
 		}
+
+		if options.OnDemandStart != nil {
+			onDemandStart := *options.OnDemandStart
+			optionsCopy.OnDemandStart = &onDemandStart
+		}
+
+		if options.IdleTimeout != nil {
+			idleTimeout := *options.IdleTimeout
+			if idleTimeout < 0 {
+				log.Printf("Instance %s IdleTimeout value (%d) cannot be negative, setting to 0 minutes", name, idleTimeout)
+				idleTimeout = 0
+			}
+			optionsCopy.IdleTimeout = &idleTimeout
+		}
 	}

 	return optionsCopy
@@ -142,6 +180,16 @@ func applyDefaultOptions(options *CreateInstanceOptions, globalSettings *config.
 		defaultRestartDelay := globalSettings.DefaultRestartDelay
 		options.RestartDelay = &defaultRestartDelay
 	}
+
+	if options.OnDemandStart == nil {
+		defaultOnDemandStart := globalSettings.DefaultOnDemandStart
+		options.OnDemandStart = &defaultOnDemandStart
+	}
+
+	if options.IdleTimeout == nil {
+		defaultIdleTimeout := 0
+		options.IdleTimeout = &defaultIdleTimeout
+	}
 }

 // NewInstance creates a new instance with the given name, log path, and options
@@ -158,10 +206,8 @@ func NewInstance(name string, globalSettings *config.InstancesConfig, options *C
 		options:        optionsCopy,
 		globalSettings: globalSettings,
 		logger:         logger,
-
-		Running: false,
-
-		Created: time.Now().Unix(),
+		timeProvider:   realTimeProvider{},
+		Created:        time.Now().Unix(),
 	}
 }

@@ -189,6 +235,11 @@ func (i *Process) SetOptions(options *CreateInstanceOptions) {
 	i.proxy = nil
 }

+// SetTimeProvider sets a custom time provider for testing
+func (i *Process) SetTimeProvider(tp TimeProvider) {
+	i.timeProvider = tp
+}
+
 // GetProxy returns the reverse proxy for this instance, creating it if needed
 func (i *Process) GetProxy() (*httputil.ReverseProxy, error) {
 	i.mu.Lock()
--- a/pkg/instance/instance_test.go
+++ b/pkg/instance/instance_test.go
@@ -91,38 +91,6 @@ func TestNewInstance_WithRestartOptions(t *testing.T) {
 	}
 }

-func TestNewInstance_ValidationAndDefaults(t *testing.T) {
-	globalSettings := &config.InstancesConfig{
-		LogsDir:             "/tmp/test",
-		DefaultAutoRestart:  true,
-		DefaultMaxRestarts:  3,
-		DefaultRestartDelay: 5,
-	}
-
-	// Test with invalid negative values
-	invalidMaxRestarts := -5
-	invalidRestartDelay := -10
-
-	options := &instance.CreateInstanceOptions{
-		MaxRestarts:  &invalidMaxRestarts,
-		RestartDelay: &invalidRestartDelay,
-		LlamaServerOptions: llamacpp.LlamaServerOptions{
-			Model: "/path/to/model.gguf",
-		},
-	}
-
-	instance := instance.NewInstance("test-instance", globalSettings, options)
-	opts := instance.GetOptions()
-
-	// Check that negative values were corrected to 0
-	if opts.MaxRestarts == nil || *opts.MaxRestarts != 0 {
-		t.Errorf("Expected MaxRestarts to be corrected to 0, got %v", opts.MaxRestarts)
-	}
-	if opts.RestartDelay == nil || *opts.RestartDelay != 0 {
-		t.Errorf("Expected RestartDelay to be corrected to 0, got %v", opts.RestartDelay)
-	}
-}
-
 func TestSetOptions(t *testing.T) {
 	globalSettings := &config.InstancesConfig{
 		LogsDir:             "/tmp/test",
@@ -164,33 +132,6 @@ func TestSetOptions(t *testing.T) {
 	}
 }

-func TestSetOptions_NilOptions(t *testing.T) {
-	globalSettings := &config.InstancesConfig{
-		LogsDir:             "/tmp/test",
-		DefaultAutoRestart:  true,
-		DefaultMaxRestarts:  3,
-		DefaultRestartDelay: 5,
-	}
-
-	options := &instance.CreateInstanceOptions{
-		LlamaServerOptions: llamacpp.LlamaServerOptions{
-			Model: "/path/to/model.gguf",
-		},
-	}
-
-	instance := instance.NewInstance("test-instance", globalSettings, options)
-	originalOptions := instance.GetOptions()
-
-	// Try to set nil options
-	instance.SetOptions(nil)
-
-	// Options should remain unchanged
-	currentOptions := instance.GetOptions()
-	if currentOptions.Model != originalOptions.Model {
-		t.Error("Options should not change when setting nil options")
-	}
-}
-
 func TestGetProxy(t *testing.T) {
 	globalSettings := &config.InstancesConfig{
 		LogsDir: "/tmp/test",
@@ -317,58 +258,6 @@ func TestUnmarshalJSON(t *testing.T) {
 	}
 }

-func TestUnmarshalJSON_PartialOptions(t *testing.T) {
-	jsonData := `{
-		"name": "test-instance",
-		"running": false,
-		"options": {
-			"model": "/path/to/model.gguf"
-		}
-	}`
-
-	var inst instance.Process
-	err := json.Unmarshal([]byte(jsonData), &inst)
-	if err != nil {
-		t.Fatalf("JSON unmarshal failed: %v", err)
-	}
-
-	opts := inst.GetOptions()
-	if opts.Model != "/path/to/model.gguf" {
-		t.Errorf("Expected model '/path/to/model.gguf', got %q", opts.Model)
-	}
-
-	// Note: Defaults are NOT applied during unmarshaling
-	// They should only be applied by NewInstance or SetOptions
-	if opts.AutoRestart != nil {
-		t.Error("Expected AutoRestart to be nil (no defaults applied during unmarshal)")
-	}
-}
-
-func TestUnmarshalJSON_NoOptions(t *testing.T) {
-	jsonData := `{
-		"name": "test-instance",
-		"running": false
-	}`
-
-	var inst instance.Process
-	err := json.Unmarshal([]byte(jsonData), &inst)
-	if err != nil {
-		t.Fatalf("JSON unmarshal failed: %v", err)
-	}
-
-	if inst.Name != "test-instance" {
-		t.Errorf("Expected name 'test-instance', got %q", inst.Name)
-	}
-	if inst.Running {
-		t.Error("Expected running to be false")
-	}
-
-	opts := inst.GetOptions()
-	if opts != nil {
-		t.Error("Expected options to be nil when not provided in JSON")
-	}
-}
-
 func TestCreateInstanceOptionsValidation(t *testing.T) {
 	tests := []struct {
 		name          string
@@ -377,13 +266,6 @@ func TestCreateInstanceOptionsValidation(t *testing.T) {
 		expectedMax   int
 		expectedDelay int
 	}{
-		{
-			name:          "nil values",
-			maxRestarts:   nil,
-			restartDelay:  nil,
-			expectedMax:   0, // Should remain nil, but we can't easily test nil in this structure
-			expectedDelay: 0,
-		},
 		{
 			name:          "valid positive values",
 			maxRestarts:   testutil.IntPtr(10),
@@ -424,20 +306,16 @@ func TestCreateInstanceOptionsValidation(t *testing.T) {
 			instance := instance.NewInstance("test", globalSettings, options)
 			opts := instance.GetOptions()

-			if tt.maxRestarts != nil {
-				if opts.MaxRestarts == nil {
-					t.Error("Expected MaxRestarts to be set")
-				} else if *opts.MaxRestarts != tt.expectedMax {
-					t.Errorf("Expected MaxRestarts %d, got %d", tt.expectedMax, *opts.MaxRestarts)
-				}
+			if opts.MaxRestarts == nil {
+				t.Error("Expected MaxRestarts to be set")
+			} else if *opts.MaxRestarts != tt.expectedMax {
+				t.Errorf("Expected MaxRestarts %d, got %d", tt.expectedMax, *opts.MaxRestarts)
 			}

-			if tt.restartDelay != nil {
-				if opts.RestartDelay == nil {
-					t.Error("Expected RestartDelay to be set")
-				} else if *opts.RestartDelay != tt.expectedDelay {
-					t.Errorf("Expected RestartDelay %d, got %d", tt.expectedDelay, *opts.RestartDelay)
-				}
+			if opts.RestartDelay == nil {
+				t.Error("Expected RestartDelay to be set")
+			} else if *opts.RestartDelay != tt.expectedDelay {
+				t.Errorf("Expected RestartDelay %d, got %d", tt.expectedDelay, *opts.RestartDelay)
 			}
 		})
 	}
--- a/pkg/instance/lifecycle.go
+++ b/pkg/instance/lifecycle.go
@@ -4,6 +4,7 @@ import (
 	"context"
 	"fmt"
 	"log"
+	"net/http"
 	"os/exec"
 	"runtime"
 	"syscall"
@@ -30,6 +31,9 @@ func (i *Process) Start() error {
 		i.restarts = 0
 	}

+	// Initialize last request time to current time when starting
+	i.lastRequestTime.Store(i.timeProvider.Now().Unix())
+
 	// Create log files
 	if err := i.logger.Create(); err != nil {
 		return fmt.Errorf("failed to create log files: %w", err)
@@ -140,6 +144,74 @@ func (i *Process) Stop() error {
 	return nil
 }

+func (i *Process) WaitForHealthy(timeout int) error {
+	if !i.Running {
+		return fmt.Errorf("instance %s is not running", i.Name)
+	}
+
+	if timeout <= 0 {
+		timeout = 30 // Default to 30 seconds if no timeout is specified
+	}
+
+	ctx, cancel := context.WithTimeout(context.Background(), time.Duration(timeout)*time.Second)
+	defer cancel()
+
+	// Get instance options to build the health check URL
+	opts := i.GetOptions()
+	if opts == nil {
+		return fmt.Errorf("instance %s has no options set", i.Name)
+	}
+
+	// Build the health check URL directly
+	host := opts.Host
+	if host == "" {
+		host = "localhost"
+	}
+	healthURL := fmt.Sprintf("http://%s:%d/health", host, opts.Port)
+
+	// Create a dedicated HTTP client for health checks
+	client := &http.Client{
+		Timeout: 5 * time.Second, // 5 second timeout per request
+	}
+
+	// Helper function to check health directly
+	checkHealth := func() bool {
+		req, err := http.NewRequestWithContext(ctx, "GET", healthURL, nil)
+		if err != nil {
+			return false
+		}
+
+		resp, err := client.Do(req)
+		if err != nil {
+			return false
+		}
+		defer resp.Body.Close()
+
+		return resp.StatusCode == http.StatusOK
+	}
+
+	// Try immediate check first
+	if checkHealth() {
+		return nil // Instance is healthy
+	}
+
+	// If immediate check failed, start polling
+	ticker := time.NewTicker(1 * time.Second)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ctx.Done():
+			return fmt.Errorf("timeout waiting for instance %s to become healthy after %d seconds", i.Name, timeout)
+		case <-ticker.C:
+			if checkHealth() {
+				return nil // Instance is healthy
+			}
+			// Continue polling
+		}
+	}
+}
+
 func (i *Process) monitorProcess() {
 	defer func() {
 		i.mu.Lock()
--- a/pkg/instance/timeout.go
+++ b/pkg/instance/timeout.go
@@ -0,0 +1,28 @@
+package instance
+
+// UpdateLastRequestTime updates the last request access time for the instance via proxy
+func (i *Process) UpdateLastRequestTime() {
+	i.mu.Lock()
+	defer i.mu.Unlock()
+
+	lastRequestTime := i.timeProvider.Now().Unix()
+	i.lastRequestTime.Store(lastRequestTime)
+}
+
+func (i *Process) ShouldTimeout() bool {
+	i.mu.RLock()
+	defer i.mu.RUnlock()
+
+	if !i.Running || i.options.IdleTimeout == nil || *i.options.IdleTimeout <= 0 {
+		return false
+	}
+
+	// Check if the last request time exceeds the idle timeout
+	lastRequest := i.lastRequestTime.Load()
+	idleTimeoutMinutes := *i.options.IdleTimeout
+
+	// Convert timeout from minutes to seconds for comparison
+	idleTimeoutSeconds := int64(idleTimeoutMinutes * 60)
+
+	return (i.timeProvider.Now().Unix() - lastRequest) > idleTimeoutSeconds
+}
--- a/pkg/instance/timeout_test.go
+++ b/pkg/instance/timeout_test.go
@@ -0,0 +1,195 @@
+package instance_test
+
+import (
+	"llamactl/pkg/backends/llamacpp"
+	"llamactl/pkg/config"
+	"llamactl/pkg/instance"
+	"llamactl/pkg/testutil"
+	"sync/atomic"
+	"testing"
+	"time"
+)
+
+// MockTimeProvider implements TimeProvider for testing
+type MockTimeProvider struct {
+	currentTime atomic.Int64 // Unix timestamp
+}
+
+func NewMockTimeProvider(t time.Time) *MockTimeProvider {
+	m := &MockTimeProvider{}
+	m.currentTime.Store(t.Unix())
+	return m
+}
+
+func (m *MockTimeProvider) Now() time.Time {
+	return time.Unix(m.currentTime.Load(), 0)
+}
+
+func (m *MockTimeProvider) SetTime(t time.Time) {
+	m.currentTime.Store(t.Unix())
+}
+
+// Timeout-related tests
+
+func TestUpdateLastRequestTime(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	options := &instance.CreateInstanceOptions{
+		LlamaServerOptions: llamacpp.LlamaServerOptions{
+			Model: "/path/to/model.gguf",
+		},
+	}
+
+	inst := instance.NewInstance("test-instance", globalSettings, options)
+
+	// Test that UpdateLastRequestTime doesn't panic
+	inst.UpdateLastRequestTime()
+}
+
+func TestShouldTimeout_NotRunning(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	idleTimeout := 1 // 1 minute
+	options := &instance.CreateInstanceOptions{
+		IdleTimeout: &idleTimeout,
+		LlamaServerOptions: llamacpp.LlamaServerOptions{
+			Model: "/path/to/model.gguf",
+		},
+	}
+
+	inst := instance.NewInstance("test-instance", globalSettings, options)
+
+	// Instance is not running, should not timeout regardless of configuration
+	if inst.ShouldTimeout() {
+		t.Error("Non-running instance should never timeout")
+	}
+}
+
+func TestShouldTimeout_NoTimeoutConfigured(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	tests := []struct {
+		name        string
+		idleTimeout *int
+	}{
+		{"nil timeout", nil},
+		{"zero timeout", testutil.IntPtr(0)},
+		{"negative timeout", testutil.IntPtr(-5)},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			options := &instance.CreateInstanceOptions{
+				IdleTimeout: tt.idleTimeout,
+				LlamaServerOptions: llamacpp.LlamaServerOptions{
+					Model: "/path/to/model.gguf",
+				},
+			}
+
+			inst := instance.NewInstance("test-instance", globalSettings, options)
+			// Simulate running state
+			inst.Running = true
+
+			if inst.ShouldTimeout() {
+				t.Errorf("Instance with %s should not timeout", tt.name)
+			}
+		})
+	}
+}
+
+func TestShouldTimeout_WithinTimeLimit(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	idleTimeout := 5 // 5 minutes
+	options := &instance.CreateInstanceOptions{
+		IdleTimeout: &idleTimeout,
+		LlamaServerOptions: llamacpp.LlamaServerOptions{
+			Model: "/path/to/model.gguf",
+		},
+	}
+
+	inst := instance.NewInstance("test-instance", globalSettings, options)
+	inst.Running = true
+
+	// Update last request time to now
+	inst.UpdateLastRequestTime()
+
+	// Should not timeout immediately
+	if inst.ShouldTimeout() {
+		t.Error("Instance should not timeout when last request was recent")
+	}
+}
+
+func TestShouldTimeout_ExceedsTimeLimit(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	idleTimeout := 1 // 1 minute
+	options := &instance.CreateInstanceOptions{
+		IdleTimeout: &idleTimeout,
+		LlamaServerOptions: llamacpp.LlamaServerOptions{
+			Model: "/path/to/model.gguf",
+		},
+	}
+
+	inst := instance.NewInstance("test-instance", globalSettings, options)
+	inst.Running = true
+
+	// Use MockTimeProvider to simulate old last request time
+	mockTime := NewMockTimeProvider(time.Now())
+	inst.SetTimeProvider(mockTime)
+
+	// Set last request time to now
+	inst.UpdateLastRequestTime()
+
+	// Advance time by 2 minutes (exceeds 1 minute timeout)
+	mockTime.SetTime(time.Now().Add(2 * time.Minute))
+
+	if !inst.ShouldTimeout() {
+		t.Error("Instance should timeout when last request exceeds idle timeout")
+	}
+}
+
+func TestTimeoutConfiguration_Validation(t *testing.T) {
+	globalSettings := &config.InstancesConfig{
+		LogsDir: "/tmp/test",
+	}
+
+	tests := []struct {
+		name            string
+		inputTimeout    *int
+		expectedTimeout int
+	}{
+		{"default value when nil", nil, 0},
+		{"positive value", testutil.IntPtr(10), 10},
+		{"zero value", testutil.IntPtr(0), 0},
+		{"negative value gets corrected", testutil.IntPtr(-5), 0},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			options := &instance.CreateInstanceOptions{
+				IdleTimeout: tt.inputTimeout,
+				LlamaServerOptions: llamacpp.LlamaServerOptions{
+					Model: "/path/to/model.gguf",
+				},
+			}
+
+			inst := instance.NewInstance("test-instance", globalSettings, options)
+			opts := inst.GetOptions()
+
+			if opts.IdleTimeout == nil || *opts.IdleTimeout != tt.expectedTimeout {
+				t.Errorf("Expected IdleTimeout %d, got %v", tt.expectedTimeout, opts.IdleTimeout)
+			}
+		})
+	}
+}
--- a/pkg/manager/manager.go
+++ b/pkg/manager/manager.go
@@ -10,6 +10,7 @@ import (
 	"path/filepath"
 	"strings"
 	"sync"
+	"time"
 )

 // InstanceManager defines the interface for managing instances of the llama server.
@@ -31,20 +32,48 @@ type instanceManager struct {
 	instances       map[string]*instance.Process
 	ports           map[int]bool
 	instancesConfig config.InstancesConfig
+
+	// Timeout checker
+	timeoutChecker *time.Ticker
+	shutdownChan   chan struct{}
+	shutdownDone   chan struct{}
+	isShutdown     bool
 }

 // NewInstanceManager creates a new instance of InstanceManager.
 func NewInstanceManager(instancesConfig config.InstancesConfig) InstanceManager {
+	if instancesConfig.TimeoutCheckInterval <= 0 {
+		instancesConfig.TimeoutCheckInterval = 5 // Default to 5 minutes if not set
+	}
 	im := &instanceManager{
 		instances:       make(map[string]*instance.Process),
 		ports:           make(map[int]bool),
 		instancesConfig: instancesConfig,
+
+		timeoutChecker: time.NewTicker(time.Duration(instancesConfig.TimeoutCheckInterval) * time.Minute),
+		shutdownChan:   make(chan struct{}),
+		shutdownDone:   make(chan struct{}),
 	}

 	// Load existing instances from disk
 	if err := im.loadInstances(); err != nil {
 		log.Printf("Error loading instances: %v", err)
 	}
+
+	// Start the timeout checker goroutine after initialization is complete
+	go func() {
+		defer close(im.shutdownDone)
+
+		for {
+			select {
+			case <-im.timeoutChecker.C:
+				im.checkAllTimeouts()
+			case <-im.shutdownChan:
+				return // Exit goroutine on shutdown
+			}
+		}
+	}()
+
 	return im
 }

@@ -94,6 +123,27 @@ func (im *instanceManager) Shutdown() {
 	im.mu.Lock()
 	defer im.mu.Unlock()

+	// Check if already shutdown
+	if im.isShutdown {
+		return
+	}
+	im.isShutdown = true
+
+	// Signal the timeout checker to stop
+	close(im.shutdownChan)
+
+	// Release lock temporarily to wait for goroutine
+	im.mu.Unlock()
+	// Wait for the timeout checker goroutine to actually stop
+	<-im.shutdownDone
+	// Reacquire lock
+	im.mu.Lock()
+
+	// Now stop the ticker
+	if im.timeoutChecker != nil {
+		im.timeoutChecker.Stop()
+	}
+
 	var wg sync.WaitGroup
 	wg.Add(len(im.instances))

--- a/pkg/manager/manager_test.go
+++ b/pkg/manager/manager_test.go
--- a/pkg/manager/operations.go
+++ b/pkg/manager/operations.go
@@ -27,10 +27,6 @@ func (im *instanceManager) CreateInstance(name string, options *instance.CreateI
 		return nil, fmt.Errorf("instance options cannot be nil")
 	}

-	if len(im.instances) >= im.instancesConfig.MaxInstances && im.instancesConfig.MaxInstances != -1 {
-		return nil, fmt.Errorf("maximum number of instances (%d) reached", im.instancesConfig.MaxInstances)
-	}
-
 	name, err := validation.ValidateInstanceName(name)
 	if err != nil {
 		return nil, err
@@ -44,6 +40,11 @@ func (im *instanceManager) CreateInstance(name string, options *instance.CreateI
 	im.mu.Lock()
 	defer im.mu.Unlock()

+	// Check max instances limit after acquiring the lock
+	if len(im.instances) >= im.instancesConfig.MaxInstances && im.instancesConfig.MaxInstances != -1 {
+		return nil, fmt.Errorf("maximum number of instances (%d) reached", im.instancesConfig.MaxInstances)
+	}
+
 	// Check if instance with this name already exists
 	if im.instances[name] != nil {
 		return nil, fmt.Errorf("instance with name %s already exists", name)
--- a/pkg/manager/timeout.go
+++ b/pkg/manager/timeout.go
@@ -0,0 +1,26 @@
+package manager
+
+import "log"
+
+func (im *instanceManager) checkAllTimeouts() {
+	im.mu.RLock()
+	var timeoutInstances []string
+
+	// Identify instances that should timeout
+	for _, inst := range im.instances {
+		if inst.ShouldTimeout() {
+			timeoutInstances = append(timeoutInstances, inst.Name)
+		}
+	}
+	im.mu.RUnlock() // Release read lock before calling StopInstance
+
+	// Stop the timed-out instances
+	for _, name := range timeoutInstances {
+		log.Printf("Instance %s has timed out, stopping it", name)
+		if _, err := im.StopInstance(name); err != nil {
+			log.Printf("Error stopping instance %s: %v", name, err)
+		} else {
+			log.Printf("Instance %s stopped successfully", name)
+		}
+	}
+}
--- a/pkg/server/handlers.go
+++ b/pkg/server/handlers.go
@@ -28,7 +28,23 @@ func NewHandler(im manager.InstanceManager, cfg config.AppConfig) *Handler {
 	}
 }

-// HelpHandler godoc
+// VersionHandler godoc
+// @Summary Get llamactl version
+// @Description Returns the version of the llamactl command
+// @Tags version
+// @Security ApiKeyAuth
+// @Produces text/plain
+// @Success 200 {string} string "Version information"
+// @Failure 500 {string} string "Internal Server Error"
+// @Router /version [get]
+func (h *Handler) VersionHandler() http.HandlerFunc {
+	return func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/plain")
+		fmt.Fprintf(w, "Version: %s\nCommit: %s\nBuild Time: %s\n", h.cfg.Version, h.cfg.CommitHash, h.cfg.BuildTime)
+	}
+}
+
+// LlamaServerHelpHandler godoc
 // @Summary Get help for llama server
 // @Description Returns the help text for the llama server command
 // @Tags server
@@ -37,7 +53,7 @@ func NewHandler(im manager.InstanceManager, cfg config.AppConfig) *Handler {
 // @Success 200 {string} string "Help text"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /server/help [get]
-func (h *Handler) HelpHandler() http.HandlerFunc {
+func (h *Handler) LlamaServerHelpHandler() http.HandlerFunc {
 	return func(w http.ResponseWriter, r *http.Request) {
 		helpCmd := exec.Command("llama-server", "--help")
 		output, err := helpCmd.CombinedOutput()
@@ -50,7 +66,7 @@ func (h *Handler) HelpHandler() http.HandlerFunc {
 	}
 }

-// VersionHandler godoc
+// LlamaServerVersionHandler godoc
 // @Summary Get version of llama server
 // @Description Returns the version of the llama server command
 // @Tags server
@@ -59,7 +75,7 @@ func (h *Handler) HelpHandler() http.HandlerFunc {
 // @Success 200 {string} string "Version information"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /server/version [get]
-func (h *Handler) VersionHandler() http.HandlerFunc {
+func (h *Handler) LlamaServerVersionHandler() http.HandlerFunc {
 	return func(w http.ResponseWriter, r *http.Request) {
 		versionCmd := exec.Command("llama-server", "--version")
 		output, err := versionCmd.CombinedOutput()
@@ -72,7 +88,7 @@ func (h *Handler) VersionHandler() http.HandlerFunc {
 	}
 }

-// ListDevicesHandler godoc
+// LlamaServerListDevicesHandler godoc
 // @Summary List available devices for llama server
 // @Description Returns a list of available devices for the llama server
 // @Tags server
@@ -81,7 +97,7 @@ func (h *Handler) VersionHandler() http.HandlerFunc {
 // @Success 200 {string} string "List of devices"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /server/devices [get]
-func (h *Handler) ListDevicesHandler() http.HandlerFunc {
+func (h *Handler) LlamaServerListDevicesHandler() http.HandlerFunc {
 	return func(w http.ResponseWriter, r *http.Request) {
 		listCmd := exec.Command("llama-server", "--list-devices")
 		output, err := listCmd.CombinedOutput()
@@ -100,7 +116,7 @@ func (h *Handler) ListDevicesHandler() http.HandlerFunc {
 // @Tags instances
 // @Security ApiKeyAuth
 // @Produces json
-// @Success 200 {array} Instance "List of instances"
+// @Success 200 {array} instance.Process "List of instances"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances [get]
 func (h *Handler) ListInstances() http.HandlerFunc {
@@ -127,8 +143,8 @@ func (h *Handler) ListInstances() http.HandlerFunc {
 // @Accept json
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Param options body CreateInstanceOptions true "Instance configuration options"
-// @Success 201 {object} Instance "Created instance details"
+// @Param options body instance.CreateInstanceOptions true "Instance configuration options"
+// @Success 201 {object} instance.Process "Created instance details"
 // @Failure 400 {string} string "Invalid request body"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name} [post]
@@ -168,7 +184,7 @@ func (h *Handler) CreateInstance() http.HandlerFunc {
 // @Security ApiKeyAuth
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Success 200 {object} Instance "Instance details"
+// @Success 200 {object} instance.Process "Instance details"
 // @Failure 400 {string} string "Invalid name format"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name} [get]
@@ -202,8 +218,8 @@ func (h *Handler) GetInstance() http.HandlerFunc {
 // @Accept json
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Param options body CreateInstanceOptions true "Instance configuration options"
-// @Success 200 {object} Instance "Updated instance details"
+// @Param options body instance.CreateInstanceOptions true "Instance configuration options"
+// @Success 200 {object} instance.Process "Updated instance details"
 // @Failure 400 {string} string "Invalid name format"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name} [put]
@@ -242,7 +258,7 @@ func (h *Handler) UpdateInstance() http.HandlerFunc {
 // @Security ApiKeyAuth
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Success 200 {object} Instance "Started instance details"
+// @Success 200 {object} instance.Process "Started instance details"
 // @Failure 400 {string} string "Invalid name format"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name}/start [post]
@@ -275,7 +291,7 @@ func (h *Handler) StartInstance() http.HandlerFunc {
 // @Security ApiKeyAuth
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Success 200 {object} Instance "Stopped instance details"
+// @Success 200 {object} instance.Process "Stopped instance details"
 // @Failure 400 {string} string "Invalid name format"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name}/stop [post]
@@ -308,7 +324,7 @@ func (h *Handler) StopInstance() http.HandlerFunc {
 // @Security ApiKeyAuth
 // @Produces json
 // @Param name path string true "Instance Name"
-// @Success 200 {object} Instance "Restarted instance details"
+// @Success 200 {object} instance.Process "Restarted instance details"
 // @Failure 400 {string} string "Invalid name format"
 // @Failure 500 {string} string "Internal Server Error"
 // @Router /instances/{name}/restart [post]
@@ -456,6 +472,9 @@ func (h *Handler) ProxyToInstance() http.HandlerFunc {
 			proxyPath = "/" + proxyPath
 		}

+		// Update the last request time for the instance
+		inst.UpdateLastRequestTime()
+
 		// Modify the request to remove the proxy prefix
 		originalPath := r.URL.Path
 		r.URL.Path = proxyPath
@@ -556,8 +575,23 @@ func (h *Handler) OpenAIProxy() http.HandlerFunc {
 		}

 		if !inst.Running {
-			http.Error(w, "Instance is not running", http.StatusServiceUnavailable)
-			return
+			if inst.GetOptions().OnDemandStart != nil && *inst.GetOptions().OnDemandStart {
+				// If on-demand start is enabled, start the instance
+				if _, err := h.InstanceManager.StartInstance(modelName); err != nil {
+					http.Error(w, "Failed to start instance: "+err.Error(), http.StatusInternalServerError)
+					return
+				}
+
+				// Wait for the instance to become healthy before proceeding
+				if err := inst.WaitForHealthy(h.cfg.Instances.OnDemandStartTimeout); err != nil { // 2 minutes timeout
+					http.Error(w, "Instance failed to become healthy: "+err.Error(), http.StatusServiceUnavailable)
+					return
+				}
+
+			} else {
+				http.Error(w, "Instance is not running", http.StatusServiceUnavailable)
+				return
+			}
 		}

 		proxy, err := inst.GetProxy()
@@ -566,6 +600,9 @@ func (h *Handler) OpenAIProxy() http.HandlerFunc {
 			return
 		}

+		// Update last request time for the instance
+		inst.UpdateLastRequestTime()
+
 		// Recreate the request body from the bytes we read
 		r.Body = io.NopCloser(bytes.NewReader(bodyBytes))
 		r.ContentLength = int64(len(bodyBytes))
--- a/pkg/server/routes.go
+++ b/pkg/server/routes.go
@@ -8,7 +8,7 @@ import (
 	"github.com/go-chi/cors"
 	httpSwagger "github.com/swaggo/http-swagger"

-	_ "llamactl/docs"
+	_ "llamactl/apidocs"
 	"llamactl/webui"
 )

@@ -42,10 +42,12 @@ func SetupRouter(handler *Handler) *chi.Mux {
 			r.Use(authMiddleware.AuthMiddleware(KeyTypeManagement))
 		}

+		r.Get("/version", handler.VersionHandler()) // Get server version
+
 		r.Route("/server", func(r chi.Router) {
-			r.Get("/help", handler.HelpHandler())
-			r.Get("/version", handler.VersionHandler())
-			r.Get("/devices", handler.ListDevicesHandler())
+			r.Get("/help", handler.LlamaServerHelpHandler())
+			r.Get("/version", handler.LlamaServerVersionHandler())
+			r.Get("/devices", handler.LlamaServerListDevicesHandler())
 		})

 		// Instance management endpoints
--- a/webui/src/App.tsx
+++ b/webui/src/App.tsx
@@ -7,6 +7,7 @@ import SystemInfoDialog from "./components/SystemInfoDialog";
 import { type CreateInstanceOptions, type Instance } from "@/types/instance";
 import { useInstances } from "@/contexts/InstancesContext";
 import { useAuth } from "@/contexts/AuthContext";
+import { ThemeProvider } from "@/contexts/ThemeContext";

 function App() {
  const { isAuthenticated, isLoading: authLoading } = useAuth();
@@ -42,44 +43,50 @@ function App() {
  // Show loading spinner while checking auth
  if (authLoading) {
    return (
-      <div className="min-h-screen bg-gray-50 flex items-center justify-center">
-        <div className="text-center">
-          <div className="animate-spin rounded-full h-8 w-8 border-b-2 border-blue-600 mx-auto mb-4"></div>
-          <p className="text-gray-600">Loading...</p>
+      <ThemeProvider>
+        <div className="min-h-screen bg-background flex items-center justify-center">
+          <div className="text-center">
+            <div className="animate-spin rounded-full h-8 w-8 border-b-2 border-primary mx-auto mb-4"></div>
+            <p className="text-muted-foreground">Loading...</p>
+          </div>
        </div>
-      </div>
+      </ThemeProvider>
    );
  }

  // Show login dialog if not authenticated
  if (!isAuthenticated) {
    return (
-      <div className="min-h-screen bg-gray-50">
-        <LoginDialog open={true} />
-      </div>
+      <ThemeProvider>
+        <div className="min-h-screen bg-background">
+          <LoginDialog open={true} />
+        </div>
+      </ThemeProvider>
    );
  }

  // Show main app if authenticated
  return (
-    <div className="min-h-screen bg-gray-50">
-      <Header onCreateInstance={handleCreateInstance} onShowSystemInfo={handleShowSystemInfo} />
-      <main className="container mx-auto max-w-4xl px-4 py-8">
-        <InstanceList editInstance={handleEditInstance} />
-      </main>
+    <ThemeProvider>
+      <div className="min-h-screen bg-background">
+        <Header onCreateInstance={handleCreateInstance} onShowSystemInfo={handleShowSystemInfo} />
+        <main className="container mx-auto max-w-4xl px-4 py-8">
+          <InstanceList editInstance={handleEditInstance} />
+        </main>

-      <InstanceDialog
-        open={isInstanceModalOpen}
-        onOpenChange={setIsInstanceModalOpen}
-        onSave={handleSaveInstance}
-        instance={editingInstance}
-      />
+        <InstanceDialog
+          open={isInstanceModalOpen}
+          onOpenChange={setIsInstanceModalOpen}
+          onSave={handleSaveInstance}
+          instance={editingInstance}
+        />

-      <SystemInfoDialog
-        open={isSystemInfoModalOpen}
-        onOpenChange={setIsSystemInfoModalOpen}
-      />
-    </div>
+        <SystemInfoDialog
+          open={isSystemInfoModalOpen}
+          onOpenChange={setIsSystemInfoModalOpen}
+        />
+      </div>
+    </ThemeProvider>
  );
 }

--- a/webui/src/tests/App.test.tsx
+++ b/webui/src/tests/App.test.tsx
@@ -55,6 +55,21 @@ describe('App Component - Critical Business Logic Only', () => {
    vi.mocked(instancesApi.list).mockResolvedValue(mockInstances)
    window.sessionStorage.setItem('llamactl_management_key', 'test-api-key-123')
    global.fetch = vi.fn(() => Promise.resolve(new Response(null, { status: 200 })))
+    
+    // Mock window.matchMedia for dark mode functionality
+    Object.defineProperty(window, 'matchMedia', {
+      writable: true,
+      value: vi.fn().mockImplementation((query: string) => ({
+        matches: false,
+        media: query,
+        onchange: null,
+        addListener: vi.fn(),
+        removeListener: vi.fn(),
+        addEventListener: vi.fn(),
+        removeEventListener: vi.fn(),
+        dispatchEvent: vi.fn(),
+      })),
+    })
  })

  afterEach(() => {
--- a/webui/src/components/Header.tsx
+++ b/webui/src/components/Header.tsx
@@ -1,6 +1,7 @@
 import { Button } from "@/components/ui/button";
-import { HelpCircle, LogOut } from "lucide-react";
+import { HelpCircle, LogOut, Moon, Sun } from "lucide-react";
 import { useAuth } from "@/contexts/AuthContext";
+import { useTheme } from "@/contexts/ThemeContext";

 interface HeaderProps {
  onCreateInstance: () => void;
@@ -9,6 +10,7 @@ interface HeaderProps {

 function Header({ onCreateInstance, onShowSystemInfo }: HeaderProps) {
  const { logout } = useAuth();
+  const { theme, toggleTheme } = useTheme();

  const handleLogout = () => {
    if (confirm("Are you sure you want to logout?")) {
@@ -17,10 +19,10 @@ function Header({ onCreateInstance, onShowSystemInfo }: HeaderProps) {
  };

  return (
-    <header className="bg-white border-b border-gray-200">
+    <header className="bg-card border-b border-border">
      <div className="container mx-auto max-w-4xl px-4 py-4">
        <div className="flex items-center justify-between">
-          <h1 className="text-2xl font-bold text-gray-900">
+          <h1 className="text-2xl font-bold text-foreground">
            Llamactl Dashboard
          </h1>

@@ -29,6 +31,16 @@ function Header({ onCreateInstance, onShowSystemInfo }: HeaderProps) {
              Create Instance
            </Button>

+            <Button
+              variant="outline"
+              size="icon"
+              onClick={toggleTheme}
+              data-testid="theme-toggle-button"
+              title={`Switch to ${theme === 'light' ? 'dark' : 'light'} mode`}
+            >
+              {theme === 'light' ? <Moon className="h-4 w-4" /> : <Sun className="h-4 w-4" />}
+            </Button>
+
            <Button
              variant="outline"
              size="icon"
--- a/webui/src/components/InstanceList.tsx
+++ b/webui/src/components/InstanceList.tsx
@@ -18,8 +18,8 @@ function InstanceList({ editInstance }: InstanceListProps) {
    return (
      <div className="flex items-center justify-center py-12" aria-label="Loading">
        <div className="text-center">
-          <div className="animate-spin rounded-full h-8 w-8 border-b-2 border-blue-600 mx-auto mb-4"></div>
-          <p className="text-gray-600">Loading instances...</p>
+          <div className="animate-spin rounded-full h-8 w-8 border-b-2 border-primary mx-auto mb-4"></div>
+          <p className="text-muted-foreground">Loading instances...</p>
        </div>
      </div>
    )
@@ -28,7 +28,7 @@ function InstanceList({ editInstance }: InstanceListProps) {
  if (error) {
    return (
      <div className="text-center py-12">
-        <div className="text-red-600 mb-4">
+        <div className="text-destructive mb-4">
          <p className="text-lg font-semibold">Error loading instances</p>
          <p className="text-sm">{error}</p>
        </div>
@@ -39,15 +39,15 @@ function InstanceList({ editInstance }: InstanceListProps) {
  if (instances.length === 0) {
    return (
      <div className="text-center py-12">
-        <p className="text-gray-600 text-lg mb-2">No instances found</p>
-        <p className="text-gray-500 text-sm">Create your first instance to get started</p>
+        <p className="text-foreground text-lg mb-2">No instances found</p>
+        <p className="text-muted-foreground text-sm">Create your first instance to get started</p>
      </div>
    )
  }

  return (
    <div className="space-y-4">
-      <h2 className="text-xl font-semibold text-gray-900 mb-6">
+      <h2 className="text-xl font-semibold text-foreground mb-6">
        Instances ({instances.length})
      </h2>
      
--- a/webui/src/components/SystemInfoDialog.tsx
+++ b/webui/src/components/SystemInfoDialog.tsx
@@ -19,6 +19,15 @@ import {
 } from 'lucide-react'
 import { serverApi } from '@/lib/api'

+// Helper to get version from environment
+const getAppVersion = (): string => {
+  try {
+    return (import.meta.env as Record<string, string>).VITE_APP_VERSION || 'unknown'
+  } catch {
+    return 'unknown'
+  }
+}
+
 interface SystemInfoModalProps {
  open: boolean
  onOpenChange: (open: boolean) => void
@@ -109,9 +118,20 @@ const SystemInfoDialog: React.FC<SystemInfoModalProps> = ({
            </div>
          ) : systemInfo ? (
            <div className="space-y-6">
-              {/* Version Section */}
+              {/* Llamactl Version Section */}
              <div className="space-y-3">
-                <h3 className="font-semibold">Version</h3>
+                <h3 className="font-semibold">Llamactl Version</h3>
+                
+                <div className="bg-gray-900 rounded-lg p-4">
+                  <pre className="text-sm text-gray-300 whitespace-pre-wrap font-mono">
+                    {getAppVersion()}
+                  </pre>
+                </div>
+              </div>
+
+              {/* Llama Server Version Section */}
+              <div className="space-y-3">
+                <h3 className="font-semibold">Llama Server Version</h3>
                
                <div className="bg-gray-900 rounded-lg p-4">
                  <div className="mb-2">
--- a/webui/src/contexts/ThemeContext.tsx
+++ b/webui/src/contexts/ThemeContext.tsx
@@ -0,0 +1,54 @@
+import { createContext, useContext, useEffect, useState, type ReactNode } from "react";
+
+type Theme = "light" | "dark";
+
+interface ThemeContextType {
+  theme: Theme;
+  toggleTheme: () => void;
+}
+
+const ThemeContext = createContext<ThemeContextType | undefined>(undefined);
+
+interface ThemeProviderProps {
+  children: ReactNode;
+}
+
+export function ThemeProvider({ children }: ThemeProviderProps) {
+  const [theme, setTheme] = useState<Theme>(() => {
+    const stored = localStorage.getItem("theme");
+    if (stored === "light" || stored === "dark") {
+      return stored;
+    }
+    return window.matchMedia("(prefers-color-scheme: dark)").matches ? "dark" : "light";
+  });
+
+  useEffect(() => {
+    const root = document.documentElement;
+    
+    if (theme === "dark") {
+      root.classList.add("dark");
+    } else {
+      root.classList.remove("dark");
+    }
+    
+    localStorage.setItem("theme", theme);
+  }, [theme]);
+
+  const toggleTheme = () => {
+    setTheme(prevTheme => prevTheme === "light" ? "dark" : "light");
+  };
+
+  return (
+    <ThemeContext.Provider value={{ theme, toggleTheme }}>
+      {children}
+    </ThemeContext.Provider>
+  );
+}
+
+export function useTheme() {
+  const context = useContext(ThemeContext);
+  if (context === undefined) {
+    throw new Error("useTheme must be used within a ThemeProvider");
+  }
+  return context;
+}
--- a/webui/src/lib/zodFormUtils.ts
+++ b/webui/src/lib/zodFormUtils.ts
@@ -21,6 +21,15 @@ export const basicFieldsConfig: Record<string, {
    placeholder: '5',
    description: 'Delay in seconds before attempting restart'
  },
+  idle_timeout: {
+    label: 'Idle Timeout (minutes)',
+    placeholder: '60',
+    description: 'Time in minutes before instance is considered idle and stopped'
+  },
+  on_demand_start: {
+    label: 'On-Demand Start',
+    description: 'Start instance upon receiving OpenAI-compatible API request'
+  },
  model: {
    label: 'Model Path',
    placeholder: '/path/to/model.gguf',
--- a/webui/src/schemas/instanceOptions.ts
+++ b/webui/src/schemas/instanceOptions.ts
@@ -6,6 +6,8 @@ export const CreateInstanceOptionsSchema = z.object({
  auto_restart: z.boolean().optional(),
  max_restarts: z.number().optional(),
  restart_delay: z.number().optional(),
+  idle_timeout: z.number().optional(),
+  on_demand_start: z.boolean().optional(),

  // Common params
  verbose_prompt: z.boolean().optional(),
--- a/webui/src/vite-env.d.ts
+++ b/webui/src/vite-env.d.ts
@@ -0,0 +1,13 @@
+/// <reference types="vite/client" />
+
+declare global {
+  interface ImportMetaEnv {
+    readonly VITE_APP_VERSION?: string
+  }
+
+  interface ImportMeta {
+    readonly env: ImportMetaEnv
+  }
+}
+
+export {}
--- a/webui/tsconfig.json
+++ b/webui/tsconfig.json
@@ -18,8 +18,9 @@
    "baseUrl": ".",
    "paths": {
      "@/*": ["./src/*"]
-    }
+    },
+    "types": ["vite/client"]
  },
-  "include": ["src"],
+  "include": ["src", "src/vite-env.d.ts"],
  "references": [{ "path": "./tsconfig.node.json" }]
 }
Author	SHA1	Message	Date
Matúš Námešný	a6e3cb4a9b	Merge pull request #23 from lordmathis/feat/start-on-request feat: On-Demand Instance Start	2025-08-20 16:04:59 +02:00
LordMathis	9181c3d7bc	Remove unused import from zodFormUtils.ts	2025-08-20 16:03:09 +02:00
LordMathis	1939b45312	Refactor WaitForHealthy method to use direct health check URL and simplify health check logic	2025-08-20 15:58:08 +02:00
LordMathis	8265a94bf7	Add on-demand start configuration to instance options and basic fields	2025-08-20 14:56:11 +02:00
LordMathis	4bc9362f7a	Add default on-demand start settings and timeout configuration to README	2025-08-20 14:41:42 +02:00
LordMathis	ddb54763f6	Add OnDemandStartTimeout configuration and update OpenAIProxy to use it	2025-08-20 14:25:43 +02:00
LordMathis	496ab3aa5d	Update README to clarify on-demand instance start feature	2025-08-20 14:22:55 +02:00
LordMathis	287a5e0817	Implement WaitForHealthy method and enhance OpenAIProxy to support on-demand instance start	2025-08-20 14:19:12 +02:00
LordMathis	7b4adfa0cd	Add DefaultOnDemandStart configuration and update instance options	2025-08-20 13:50:43 +02:00
Matúš Námešný	651c8b9b2c	Merge pull request #22 from lordmathis/feat/timeout feat: Implement idle instance timeout	2025-08-20 13:34:38 +02:00
LordMathis	7194e1fdd1	Update README to clarify idle timeout management and state persistence features	2025-08-20 13:32:03 +02:00
LordMathis	492c3ff270	Remove redundant timeout tests and improve test coverage for instance timeout validation	2025-08-20 13:25:56 +02:00
LordMathis	00a3cba717	Enhance shutdown handling in InstanceManager with proper synchronization and max instances check	2025-08-19 22:34:48 +02:00
LordMathis	eb1d4ab55f	Enhance timeout functionality tests to validate configuration and logic without starting instances	2025-08-19 20:52:59 +02:00
LordMathis	a9e3801eae	Refactor logging in checkAllTimeouts	2025-08-19 19:25:15 +02:00
LordMathis	1aaab96cec	Add idle timeout configuration to instance options and basic fields	2025-08-19 19:24:54 +02:00
LordMathis	78eda77e44	Enhance timeout handling in InstanceManager with goroutine recovery and shutdown support	2025-08-17 22:49:28 +02:00
LordMathis	d70bb634cd	Implement instance tests for timeout	2025-08-17 21:50:16 +02:00
LordMathis	41eaebc927	Add TimeoutCheckInterval to instance configuration in tests	2025-08-17 21:42:52 +02:00
LordMathis	c45fa13206	Initialize last request time on instance start and update timeout handling logic	2025-08-17 21:15:28 +02:00
LordMathis	5e3a28398d	Implement periodic timeout checking for instances	2025-08-17 21:10:48 +02:00
LordMathis	c734bcae4a	Move UpdateLastRequestTime method to timeout.go and add ShouldTimeout method for idle timeout handling	2025-08-17 20:37:20 +02:00
LordMathis	e4e7a82294	Implement last request time tracking for instance management	2025-08-17 19:44:57 +02:00
LordMathis	ccffbca6b2	Add timeout check interval and update instance configuration	2025-08-17 19:26:21 +02:00
LordMathis	902be409d5	Add IdleTimeout option to CreateInstanceOptions and update JSON handling	2025-08-17 19:06:09 +02:00
Matúš Námešný	eb9599f26a	Merge pull request #21 from lordmathis/feat/dark-mode feat: Implement dark theme and theme switching	2025-08-11 17:56:16 +02:00
LordMathis	ebf8dfdeab	Mock window.matchMedia for dark mode functionality in tests	2025-08-11 17:54:04 +02:00
LordMathis	f15c0840c4	Implement dark theme and theme switching	2025-08-11 17:39:56 +02:00
Matúš Námešný	e702bcb694	Create CNAME	2025-08-08 13:41:58 +02:00
Matúš Námešný	4895fbff15	Merge pull request #20 from lordmathis/docs/contributing docs: Add CONTRIBUTING.md to outline development setup and contribution process	2025-08-07 21:13:01 +02:00
LordMathis	282fe67355	Add CONTRIBUTING.md to outline development setup and contribution process	2025-08-07 21:10:01 +02:00
Matúš Námešný	96a36e1119	Merge pull request #19 from lordmathis/docs/readme-screenshot docs: Add dashboard screenshot to README	2025-08-07 19:55:15 +02:00
LordMathis	759fc58326	Update README to include dashboard screenshot	2025-08-07 19:51:34 +02:00
LordMathis	afef3d0180	Update import path for API documentation to use apidocs	2025-08-07 19:48:28 +02:00
LordMathis	a87652937f	Move swagger documentation to apidoc	2025-08-07 19:48:03 +02:00
Matúš Námešný	7bde12db47	Merge pull request #18 from lordmathis/feat/show-version feat: Show app version on backend and frontend	2025-08-07 19:11:58 +02:00
LordMathis	e2b64620b5	Expose version endpoint	2025-08-07 19:10:06 +02:00
LordMathis	3ba62af01a	Add VITE_APP_VERSION to environment and update SystemInfoDialog to display version	2025-08-07 19:01:31 +02:00
LordMathis	0150429e82	Add commit hash and build time to version output	2025-08-07 18:48:35 +02:00
LordMathis	2ecf096024	Add version flag to display llamactl version	2025-08-07 18:46:49 +02:00