Version 1.4.0

2026-01-05 01:48:23 +00:00
parent dc7a459ebb
commit 6947819742
4 changed files with 551 additions and 85 deletions
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,14 +1,15 @@
 # IGNY8 Change Log

-**Current Version:** 1.3.2  
-**Last Updated:** January 3, 2026
+**Current Version:** 1.4.0  
+**Last Updated:** January 5, 2026

 ---
-Restored
+
 ## Version History

 | Version | Date | Summary |
 |---------|------|---------|
+| 1.4.0 | Jan 5, 2026 | **Major** - AI Model Architecture Overhaul, IntegrationProvider Model, AIModelConfig with Credit System, SystemAISettings, Django Admin Reorganization |
 | 1.3.2 | Jan 3, 2026 | **Major** - Publishing Scheduler, Onboarding Wizard, Content Calendar, Design System Consolidation, Site Dashboard redesign |
 | 1.3.1 | Jan 2, 2026 | **Design System Consolidation** - Updated .rules with comprehensive design system & component rules, ESLint enforcement |
 | 1.3.0 | Jan 1, 2026 | **Major** - Automation overhaul, AI provider integrations (Anthropic/Bria), Model Registry, Global Progress tracking, CSS globalization |
@@ -34,6 +35,174 @@ Restored

 ---

+## v1.4.0 - January 5, 2026
+
+### Major Release: AI Architecture Overhaul & Model Configuration System
+
+This release introduces a comprehensive refactoring of the AI system architecture, centralizing all AI model configurations, provider API keys, and credit-based billing into a unified, database-driven system. Key improvements include the new `IntegrationProvider`, `AIModelConfig`, and `SystemAISettings` models, along with Django Admin reorganization and enhanced frontend settings pages.
+
+---
+
+### 🤖 AI Model Architecture Overhaul (MAJOR)
+
+**New Model: IntegrationProvider** (`modules/system/models.py`)
+- Centralized storage for ALL external service API keys
+- Supports providers: `openai`, `anthropic`, `runware`, `stripe`, `paypal`, `resend`
+- Fields: `provider_id` (PK), `display_name`, `provider_type` (ai/payment/email/storage)
+- API key management: `api_key`, `api_secret`, `webhook_secret`, `api_endpoint`
+- Configuration: `config` (JSONField), `is_active`, `is_sandbox`
+- Audit trail: `updated_by`, `created_at`, `updated_at`
+
+**New Model: AIModelConfig** (`business/billing/models.py`)
+- Single Source of Truth for all AI models (text + image) with pricing
+- Fields: `model_name` (unique), `model_type` (text/image), `provider`, `display_name`
+- Status: `is_default` (one per type), `is_active`
+- Text Model Pricing: `cost_per_1k_input`, `cost_per_1k_output`, `tokens_per_credit`
+- Image Model Pricing: `credits_per_image`, `quality_tier` (basic/quality/premium)
+- Model Limits: `max_tokens`, `context_window`, `capabilities` (JSONField)
+- History tracking via `simple_history`
+
+**New Model: SystemAISettings** (`modules/system/ai_settings.py`)
+- System-wide AI defaults (Singleton pattern, pk=1)
+- AI Parameters: `temperature` (0.0-2.0), `max_tokens`
+- Image Settings: `image_style`, `image_quality`, `max_images_per_article`, `image_size`
+- Helper methods: `get_effective_*` for account-specific overrides via `AccountSettings`
+- Image Style Choices: photorealistic, illustration, 3d_render, minimal_flat, artistic, cartoon
+
+**Model Registry Service** (`ai/model_registry.py`)
+- Central registry with database-driven model configuration
+- Caching layer for performance (5-minute TTL)
+- Methods: `get_model()`, `get_rate()`, `calculate_cost()`, `get_default_model()`
+- Provider methods: `get_provider()`, `get_api_key()`, `get_api_secret()`, `get_webhook_secret()`
+- List and validation: `list_models()`, `validate_model()`, `clear_cache()`
+
+---
+
+### 💰 Credit System Integration
+
+**AIModelConfig Credit Fields:**
+- `tokens_per_credit` - Text models: tokens per 1 credit (e.g., 1000 for GPT-4, 10000 for GPT-4o-mini)
+- `credits_per_image` - Image models: credits per image (1=basic, 5=quality, 15=premium)
+- `quality_tier` - Image quality tier for UI display
+
+**Credit Cost Configuration:**
+- Per `final-model-schemas.md`: Simplified to `operation_type`, `display_name`, `base_credits`, `is_active`
+- Removed `tokens_per_credit` from CreditCostConfig (moved to AIModelConfig)
+
+**New Migrations:**
+- `0025_add_aimodel_credit_fields.py` - Add credit fields to AIModelConfig
+- `0026_populate_aimodel_credits.py` - Populate default credit values
+- `0027_model_schema_update.py` - Schema updates for final model architecture
+
+---
+
+### 🔧 Django Admin Reorganization
+
+**Admin Group Structure:**
+- Consolidated admin panels with logical grouping
+- AI Configuration: AIModelConfig, IntegrationProvider, SystemAISettings
+- Billing: CreditCostConfig, BillingConfiguration, Invoices, Payments, Credit Packages
+- Custom logo support via `staticfiles/admin/img/`
+
+**AIModelConfig Admin:**
+- List display: model_name, model_type, provider, is_default, is_active
+- Bulk actions: activate/deactivate, set as default
+- Fieldsets: Basic Info, Pricing (Text), Pricing (Image), Model Limits, Capabilities
+- Import/Export support
+
+**IntegrationProvider Admin:**
+- List display: provider_id, display_name, provider_type, is_active, is_sandbox
+- Secure API key display (masked in list view)
+- Fieldsets: Provider Info, Authentication, Configuration, Status
+
+---
+
+### 🎨 Frontend Settings Updates
+
+**Site Settings Page Refactoring** (`pages/Sites/Settings.tsx`)
+- Tab restructure: General → AI Settings → Integrations → Publishing → Content Types
+- Image settings merged into AI Settings tab (previously separate)
+- Quality tier picker with credit display
+- Temperature and max_tokens controls
+- Image style selection with previews
+
+**AI Settings Tab Features:**
+- Content Generation Settings: append_to_prompt, default_tone, default_length
+- AI Parameters: temperature slider (0.0-2.0), max_tokens dropdown
+- Image Generation: quality tier selector (Basic/Quality/Premium with credit costs)
+- Image style dropdown with 6 style options
+- Max images per article control (1-8)
+
+**Removed Pages:**
+- `frontend/src/pages/Settings/AI.tsx` - Merged into Site Settings AI tab
+
+---
+
+### 📊 System Migrations
+
+**System Module Migrations:**
+- `0015_add_integration_provider.py` - Create IntegrationProvider model
+- `0016_populate_integration_providers.py` - Seed default providers (openai, runware, anthropic, stripe, paypal, resend)
+- `0017_create_ai_settings.py` - Create AISettings model
+- `0018_create_ai_settings_table.py` - Create SystemAISettings table
+- `0019_model_schema_update.py` - Final schema updates
+
+**Billing Module Migrations:**
+- `0024_update_image_models_v2.py` - Update image model configurations
+- `0025_add_aimodel_credit_fields.py` - Add credit fields
+- `0026_populate_aimodel_credits.py` - Populate credit values
+- `0027_model_schema_update.py` - Schema finalization
+
+---
+
+### 🔄 AI Core Updates
+
+**AI Core Service** (`ai/ai_core.py`)
+- Refactored to use ModelRegistry for model configuration
+- Dynamic API key retrieval from IntegrationProvider
+- Improved error handling and logging
+- Cost calculation using AIModelConfig rates
+
+**AI Tasks** (`ai/tasks.py`)
+- Updated to use new model configuration system
+- Credit deduction via AIModelConfig settings
+- Quality tier support for image generation
+
+**AI Settings** (`ai/settings.py`)
+- Simplified to use SystemAISettings singleton
+- Removed hardcoded API keys (now in IntegrationProvider)
+- Helper functions for getting effective settings with account overrides
+
+**Generate Images** (`ai/functions/generate_images.py`)
+- Quality tier-based model selection
+- Credits deducted based on `credits_per_image` from AIModelConfig
+- Style parameter support from SystemAISettings
+
+---
+
+### 📝 Documentation Updates
+
+**New Documentation:**
+- `docs/plans/4th-jan-refactor/final-model-schemas.md` - Complete model schema specification
+- `docs/plans/4th-jan-refactor/django-plan.md` - Django admin reorganization plan
+
+**Model Architecture:**
+- Total Models: 5 (IntegrationProvider, AIModelConfig, AISettings, AccountSettings, CreditCostConfig)
+- Removed: IntegrationSettings merged into AISettings/AccountSettings
+- Simplified: CreditCostConfig.tokens_per_credit moved to AIModelConfig
+
+---
+
+### 🐛 Bug Fixes & Improvements
+
+- Fixed image generation model selection for different quality tiers
+- Improved API key security (no longer stored in GlobalIntegrationSettings)
+- Fixed admin panel navigation for billing and AI configuration
+- Corrected credit calculation for token-based text models
+- Fixed singleton pattern enforcement for SystemAISettings
+
+---
+
 ## v1.3.2 - January 3, 2026

 ### Major Release: Publishing Scheduler, Onboarding Wizard & Content Calendar
--- a/docs/10-MODULES/BILLING.md
+++ b/docs/10-MODULES/BILLING.md
@@ -1,6 +1,6 @@
 # Billing Module

-**Last Verified:** December 25, 2025  
+**Last Verified:** January 5, 2026  
 **Status:** ✅ Active  
 **Backend Path:** `backend/igny8_core/modules/billing/` + `backend/igny8_core/business/billing/`  
 **Frontend Path:** `frontend/src/pages/Billing/` + `frontend/src/pages/Account/`
@@ -11,7 +11,7 @@

 | What | File | Key Items |
 |------|------|-----------|
-| Models | `business/billing/models.py` | `CreditTransaction`, `CreditUsageLog`, `CreditCostConfig` |
+| Models | `business/billing/models.py` | `CreditTransaction`, `CreditUsageLog`, `CreditCostConfig`, `AIModelConfig` |
 | Service | `business/billing/services/credit_service.py` | `CreditService` |
 | Views | `modules/billing/views.py` | `CreditBalanceViewSet`, `CreditUsageViewSet` |
 | Frontend | `pages/Account/PlansAndBillingPage.tsx` | Plans, credits, billing history |
@@ -23,6 +23,7 @@

 The Billing module manages:
 - Credit balance and transactions
+- AI model pricing and credit configuration (v1.4.0)
 - Usage tracking and limits
 - Plan enforcement
 - Payment processing
@@ -31,6 +32,37 @@ The Billing module manages:

 ## Data Models

+### AIModelConfig (NEW v1.4.0)
+
+Single Source of Truth for all AI models with pricing.
+
+| Field | Type | Purpose |
+|-------|------|---------|
+| model_name | CharField(100) | Model identifier (gpt-4o-mini, dall-e-3) |
+| model_type | CharField(20) | text / image |
+| provider | CharField(50) | openai, anthropic, runware |
+| display_name | CharField(200) | Human-readable name |
+| is_default | BooleanField | One default per type |
+| is_active | BooleanField | Enable/disable |
+| cost_per_1k_input | DecimalField | USD cost per 1K input tokens (text) |
+| cost_per_1k_output | DecimalField | USD cost per 1K output tokens (text) |
+| tokens_per_credit | IntegerField | Text: tokens per 1 credit (e.g., 1000) |
+| credits_per_image | IntegerField | Image: credits per image (1, 5, 15) |
+| quality_tier | CharField(20) | basic / quality / premium |
+| max_tokens | IntegerField | Model token limit |
+| context_window | IntegerField | Model context size |
+| capabilities | JSONField | vision, function_calling, etc. |
+
+**Credit Examples:**
+
+| Model | tokens_per_credit | credits_per_image | quality_tier |
+|-------|-------------------|-------------------|--------------|
+| gpt-4o | 1000 | - | - |
+| gpt-4o-mini | 10000 | - | - |
+| runware:97@1 | - | 1 | basic |
+| dall-e-3 | - | 5 | quality |
+| google:4@2 | - | 15 | premium |
+
 ### CreditTransaction (Ledger)

 | Field | Type | Purpose |
@@ -41,7 +73,8 @@ The Billing module manages:
 | balance_after | Decimal | Balance after transaction |
 | description | CharField | Transaction description |
 | metadata | JSON | Additional data |
-| reference_id | CharField | External reference |
+| payment | FK(Payment) | Payment that triggered this (v1.4.0) |
+| reference_id | CharField | DEPRECATED: Use payment FK |
 | created_at | DateTime | Transaction time |

 ### CreditUsageLog (Analytics)
@@ -60,24 +93,23 @@ The Billing module manages:
 | metadata | JSON | Additional data |
 | created_at | DateTime | Usage time |

-### CreditCostConfig
+### CreditCostConfig (Updated v1.4.0)

 | Field | Type | Purpose |
 |-------|------|---------|
-| operation_type | CharField | Operation name |
-| credits_cost | Decimal | Credits per unit |
-| unit | CharField | per_request/per_100_words/per_200_words/per_item/per_image |
-| display_name | CharField | UI display name |
-| description | TextField | Description |
-| is_active | Boolean | Enable/disable |
-| previous_cost | Decimal | Previous cost (audit) |
-| updated_by | FK | Last updater |
+| operation_type | CharField(50) PK | Unique operation ID |
+| display_name | CharField(100) | Human-readable name |
+| base_credits | IntegerField | Fixed credits per operation |
+| is_active | BooleanField | Enable/disable |
+| description | TextField | Admin notes |
+
+**Note:** `tokens_per_credit` moved to AIModelConfig in v1.4.0.

 ---

 ## Credit Calculation Methods

-### Method 1: Token-Based (Text Models)
+### Method 1: Token-Based (Text Models) - Updated v1.4.0

 Used for: Clustering, Ideas, Content Generation, Optimization

@@ -85,19 +117,30 @@ Used for: Clustering, Ideas, Content Generation, Optimization
 1. AI call completes
 2. OpenAI returns actual token usage
 3. Calculate: `total_tokens = input_tokens + output_tokens`
-4. Look up `CreditCostConfig` for operation
-5. Apply formula based on unit:
-   - `per_100_words`: `credits = ceil(words / 100) * cost`
-   - `per_item`: `credits = items * cost`
-6. Apply minimum credits
-7. Deduct from balance
+4. Look up `AIModelConfig.tokens_per_credit` for the model used
+5. Calculate: `credits = ceil(total_tokens / tokens_per_credit)`
+6. Deduct from balance

-### Method 2: Fixed Cost (Image Models)
+**Example:**
+- Model: `gpt-4o-mini` (tokens_per_credit = 10000)
+- Tokens used: 15000
+- Credits: `ceil(15000 / 10000)` = 2 credits
+
+### Method 2: Fixed Cost (Image Models) - Updated v1.4.0

 Used for: Image Generation

 **Flow:**
-1. Before AI call
+1. User selects quality tier (basic/quality/premium)
+2. Look up `AIModelConfig.credits_per_image` for selected tier
+3. Check balance sufficient: `balance >= num_images * credits_per_image`
+4. Deduct credits
+5. Make AI call
+
+**Example:**
+- Quality Tier: "quality" (dall-e-3, credits_per_image = 5)
+- Images: 3
+- Credits: 3 × 5 = 15 credits
 2. Calculate: `credits = num_images * cost_per_image`
 3. Check balance sufficient
 4. Deduct credits
@@ -188,18 +231,31 @@ Displays:

 ---

-## Credit Costs Reference
+## Credit Costs Reference (Updated v1.4.0)

-Current costs (from `CreditCostConfig`):
+**Text Model Credits** (from `AIModelConfig.tokens_per_credit`):

-| Operation | Unit | Default Cost |
-|-----------|------|--------------|
-| Clustering | per_request | 10 credits |
-| Idea Generation | per_item | 2 credits |
-| Content Generation | per_100_words | 5 credits |
-| Image Generation (Basic) | per_image | 3 credits |
-| Image Generation (Premium) | per_image | 5 credits |
-| Content Optimization | per_200_words | 3 credits |
+| Model | tokens_per_credit | Cost/Credit | Notes |
+|-------|-------------------|-------------|-------|
+| gpt-4o | 1000 | ~$0.015 | High quality, lower throughput |
+| gpt-4o-mini | 10000 | ~$0.001 | Fast, cost-effective |
+| gpt-4.5-preview | 500 | ~$0.05 | Highest quality |
+
+**Image Model Credits** (from `AIModelConfig.credits_per_image`):
+
+| Quality Tier | credits_per_image | Model Example | Notes |
+|--------------|-------------------|---------------|-------|
+| Basic | 1 | runware:97@1 | Fast generation |
+| Quality | 5 | dall-e-3 | Balanced |
+| Premium | 15 | google:4@2 | Best quality |
+
+**Operation Base Costs** (from `CreditCostConfig.base_credits`):
+
+| Operation | Base Credits | Notes |
+|-----------|--------------|-------|
+| Clustering | 10 | Per clustering request |
+| Idea Generation | 2 | Per idea generated |
+| Content Optimization | 5 | Per optimization run |

 ---

@@ -248,7 +304,4 @@ Current costs (from `CreditCostConfig`):

 | Feature | Status | Description |
 |---------|--------|-------------|
-| AI Model Config database | 🔜 Planned | Move model pricing to database |
-| Budget alerts | 🔜 Planned | Email when approaching limits |
-| Usage forecasting | 🔜 Planned | Predict credit usage |
-| Overage billing | 🔜 Planned | Charge for overages instead of blocking |
+| ~~AI Model Config database~~ | ✅ v1.4.0 | Model pricing moved to AIModelConfig |
--- a/docs/90-REFERENCE/AI-FUNCTIONS.md
+++ b/docs/90-REFERENCE/AI-FUNCTIONS.md
@@ -1,6 +1,6 @@
 # AI Functions Reference

-**Last Verified:** December 25, 2025
+**Last Verified:** January 5, 2026

 ---

@@ -8,13 +8,20 @@

 IGNY8's AI engine provides functions for content planning and generation. Located in `backend/igny8_core/ai/`.

-**Providers:**
- **OpenAI** - GPT-4 for text, DALL-E 3 for images
- **Runware** - Alternative image generation
+**Providers (v1.4.0):**
+- **OpenAI** - GPT-4 for text, DALL-E 3 for images (via `IntegrationProvider`)
+- **Anthropic** - Claude models for text (via `IntegrationProvider`)
+- **Runware** - Alternative image generation (via `IntegrationProvider`)
+
+**New in v1.4.0:**
+- Provider API keys stored in `IntegrationProvider` model
+- Model configurations stored in `AIModelConfig` model
+- System defaults stored in `SystemAISettings` singleton
+- Credit-based pricing per model via `AIModelConfig.tokens_per_credit` / `credits_per_image`

 ---

-## Architecture
+## Architecture (Updated v1.4.0)

 ```
 ┌─────────────────────────────────────────────────────────────────┐
@@ -25,12 +32,20 @@ IGNY8's AI engine provides functions for content planning and generation. Locate
 │  │   AIEngine   │───►│   Function   │───►│   Provider   │      │
 │  │   (router)   │    │   (logic)    │    │   (API)      │      │
 │  └──────────────┘    └──────────────┘    └──────────────┘      │
+│         │                                        ▲              │
+│         │                                        │              │
+│         ▼                                        │              │
+│  ┌──────────────┐    ┌──────────────┐    ┌──────────────┐      │
+│  │ModelRegistry │───►│AIModelConfig │───►│Integration   │      │
+│  │  (service)   │    │  (database)  │    │  Provider    │      │
+│  └──────────────┘    └──────────────┘    └──────────────┘      │
 │                                                                  │
 │  Functions:                                                      │
 │  • AutoClusterKeywords                                          │
 │  • GenerateContentIdeas                                         │
 │  • GenerateContent                                              │
 │  • GenerateImages                                               │
+│  • GenerateImagePrompts                                         │
 │  • OptimizeContent (pending)                                    │
 │                                                                  │
 └─────────────────────────────────────────────────────────────────┘
@@ -38,6 +53,38 @@ IGNY8's AI engine provides functions for content planning and generation. Locate

 ---

+## Model Registry (NEW v1.4.0)
+
+**Location:** `backend/igny8_core/ai/model_registry.py`
+
+Central registry for AI model configurations with caching.
+
+```python
+from igny8_core.ai.model_registry import ModelRegistry
+
+# Get model configuration
+model = ModelRegistry.get_model('gpt-4o-mini')
+
+# Get pricing rate
+input_rate = ModelRegistry.get_rate('gpt-4o-mini', 'input')
+
+# Calculate cost
+cost = ModelRegistry.calculate_cost('gpt-4o-mini', input_tokens=1000, output_tokens=500)
+
+# Get API key for provider
+api_key = ModelRegistry.get_api_key('openai')
+
+# Get default model
+default_text = ModelRegistry.get_default_model('text')
+default_image = ModelRegistry.get_default_model('image')
+
+# List available models
+text_models = ModelRegistry.list_models(model_type='text')
+image_models = ModelRegistry.list_models(model_type='image', provider='runware')
+```
+
+---
+
 ## AIEngine

 **Location:** `backend/igny8_core/ai/engine.py`
@@ -48,9 +95,9 @@ Main orchestrator for AI operations.
 class AIEngine:
    def __init__(self, account: Account):
        self.account = account
-        self.settings = self._load_settings()
-        self.text_provider = self._init_text_provider()
-        self.image_provider = self._init_image_provider()
+        self.settings = self._load_settings()  # Uses SystemAISettings + AccountSettings
+        self.text_provider = self._init_text_provider()  # Uses ModelRegistry
+        self.image_provider = self._init_image_provider()  # Uses ModelRegistry
    
    def auto_cluster(self, keywords: List[Keyword]) -> List[Cluster]:
        """Cluster keywords by topic"""
@@ -61,8 +108,8 @@ class AIEngine:
    def generate_content(self, task: Task) -> Content:
        """Generate full article content"""
        
-    def generate_images(self, content: Content, count: int = 1) -> List[ContentImage]:
-        """Generate images for content"""
+    def generate_images(self, content: Content, count: int = 1, quality_tier: str = 'quality') -> List[ContentImage]:
+        """Generate images for content with quality tier selection (v1.4.0)"""
 ```

 ---
@@ -103,9 +150,9 @@ class AIEngine:

 **Prompt Template:** `auto_cluster`

-**Model:** GPT-4
+**Model:** Default text model (from `AIModelConfig.is_default`)

-**Credit Cost:** 1 idea credit per operation
+**Credit Cost:** Based on `AIModelConfig.tokens_per_credit` for model used

 ---

@@ -146,9 +193,9 @@ class AIEngine:

 **Prompt Template:** `generate_ideas`

-**Model:** GPT-4
+**Model:** Default text model (from `AIModelConfig.is_default`)

-**Credit Cost:** 1 idea credit per idea generated
+**Credit Cost:** Based on `AIModelConfig.tokens_per_credit` for model used

 ---

@@ -174,7 +221,9 @@ class AIEngine:
        "target_word_count": 2000,
        "include_headings": True,
        "include_lists": True,
-        "include_code_blocks": True
+        "include_code_blocks": True,
+        "temperature": 0.7,  # From SystemAISettings or AccountSettings override
+        "max_tokens": 8192   # From SystemAISettings or AccountSettings override
    }
 }
 ```
@@ -193,15 +242,15 @@ class AIEngine:

 **Prompt Template:** `generate_content`

-**Model:** GPT-4
+**Model:** Default text model (from `AIModelConfig.is_default`)

-**Credit Cost:** 1 content credit per generation
+**Credit Cost:** Based on `AIModelConfig.tokens_per_credit` for model used

 ---

-## Function: GenerateImages
+## Function: GenerateImages (Updated v1.4.0)

-**Purpose:** Create images for article content
+**Purpose:** Create images for article content with quality tier selection

 **Input:**
 ```python
@@ -213,17 +262,27 @@ class AIEngine:
    },
    "options": {
        "count": 3,
-        "style": "photorealistic",  # photorealistic, illustration, diagram
-        "size": "1024x1024"
+        "quality_tier": "quality",  # basic (1 credit), quality (5 credits), premium (15 credits)
+        "style": "photorealistic",  # From SystemAISettings or override
+        "size": "1024x1024"  # From SystemAISettings or override
    }
 }
 ```

+**Quality Tiers (v1.4.0):**
+
+| Tier | Model Example | credits_per_image | Description |
+|------|---------------|-------------------|-------------|
+| basic | runware:97@1 | 1 | Fast generation, simple images |
+| quality | dall-e-3 | 5 | Balanced quality and speed |
+| premium | google:4@2 | 15 | Best quality, slower |
+
 **Process:**
-1. Analyze content to identify image opportunities
-2. Generate prompts for each image
-3. Call image provider API
-4. Store images and generate thumbnails
+1. Get quality tier model from `AIModelConfig.get_image_models_by_tier()`
+2. Analyze content to identify image opportunities
+3. Generate prompts for each image (via `GenerateImagePrompts`)
+4. Call image provider API (using `IntegrationProvider.get_api_key()`)
+5. Store images and generate thumbnails

 **Output:**
 ```python
@@ -235,7 +294,10 @@ class AIEngine:
            "alt_text": "Python code example showing...",
            "caption": "Example of Python list comprehension",
            "prompt": "A clean code editor showing Python syntax...",
-            "is_featured": True
+            "is_featured": True,
+            "model_used": "dall-e-3",
+            "quality_tier": "quality",
+            "credits_used": 5
        },
        ...
    ]
@@ -244,9 +306,9 @@ class AIEngine:

 **Prompt Template:** `generate_image_prompts`

-**Model:** DALL-E 3 or Runware
+**Model:** Selected by quality tier from `AIModelConfig`

-**Credit Cost:** 1 image credit per image
+**Credit Cost:** `AIModelConfig.credits_per_image` for selected model

 ---

--- a/docs/90-REFERENCE/MODELS.md
+++ b/docs/90-REFERENCE/MODELS.md
@@ -1,6 +1,6 @@
 # Database Models Reference

-**Last Verified:** January 3, 2026
+**Last Verified:** January 5, 2026

 ---

@@ -8,37 +8,219 @@

 | Scope | Models | Base Class | Filter By |
 |-------|--------|------------|-----------|
-| **Global** | `GlobalIntegrationSettings`, `GlobalAIPrompt`, `GlobalAuthorProfile`, `GlobalStrategy`, `GlobalModuleSettings`, `Industry`, `SeedKeyword` | `models.Model` | None (platform-wide) |
-| **Account** | `Account`, `User`, `Plan`, `IntegrationSettings`, `ModuleEnableSettings`, `AISettings`, `AIPrompt`, `AuthorProfile`, `CreditBalance` | `AccountBaseModel` | `account` |
+| **Global** | `IntegrationProvider`, `AIModelConfig`, `SystemAISettings`, `GlobalAIPrompt`, `GlobalAuthorProfile`, `GlobalStrategy`, `GlobalModuleSettings`, `Industry`, `SeedKeyword` | `models.Model` | None (platform-wide) |
+| **Account** | `Account`, `User`, `Plan`, `AccountSettings`, `ModuleEnableSettings`, `AISettings`, `AIPrompt`, `AuthorProfile`, `CreditBalance` | `AccountBaseModel` | `account` |
 | **Site** | `Site`, `PublishingSettings`, `AutomationConfig`, `SiteIntegration` | `AccountBaseModel` | `account`, `site` |
 | **Site+Sector** | `Keywords`, `Clusters`, `ContentIdeas`, `Tasks`, `Content`, `Images` | `SiteSectorBaseModel` | `site`, `sector` |

 ---

-## Global Models (`igny8_core/modules/system/global_settings_models.py`)
+## System Models (v1.4.0) (`igny8_core/modules/system/`)

-**Purpose:** Platform-wide defaults and API keys. Admin-only. NOT account-scoped.
+**Purpose:** Centralized AI configuration, provider API keys, and system-wide defaults.

-### GlobalIntegrationSettings (Singleton)
+### IntegrationProvider (NEW v1.4.0)
+
+Centralized storage for ALL external service API keys. Admin-only.

 ```python
-class GlobalIntegrationSettings(models.Model):
-    # OpenAI (used by ALL accounts)
-    openai_api_key = CharField(max_length=500)
-    openai_model = CharField(default='gpt-4o-mini')
-    openai_temperature = FloatField(default=0.7)
-    openai_max_tokens = IntegerField(default=8192)
+class IntegrationProvider(models.Model):
+    """Per final-model-schemas.md - Centralized API key storage"""
+    provider_id = CharField(max_length=50, primary_key=True)  # openai, runware, stripe, etc.
+    display_name = CharField(max_length=100)
+    provider_type = CharField(max_length=20)  # ai, payment, email, storage
    
-    # DALL-E / Image Generation
-    dalle_api_key = CharField(max_length=500)
-    dalle_model = CharField(default='dall-e-3')
-    dalle_size = CharField(default='1024x1024')
+    # Authentication
+    api_key = CharField(max_length=500, blank=True)
+    api_secret = CharField(max_length=500, blank=True)
+    webhook_secret = CharField(max_length=500, blank=True)
+    api_endpoint = URLField(blank=True)
    
-    # Runware
-    runware_api_key = CharField(max_length=500)
+    # Configuration
+    config = JSONField(default=dict)
+    is_active = BooleanField(default=True)
+    is_sandbox = BooleanField(default=False)
+    
+    # Audit
+    updated_by = ForeignKey(User, null=True)
+    created_at = DateTimeField(auto_now_add=True)
+    updated_at = DateTimeField(auto_now=True)
 ```

-**Critical:** Singleton (pk=1). API keys here are used by ALL accounts.
+**Seeded Providers:**
+- `openai` - AI (text + DALL-E)
+- `runware` - AI (images)
+- `anthropic` - AI (future)
+- `stripe` - Payment
+- `paypal` - Payment
+- `resend` - Email
+
+**Helper Methods:**
+- `IntegrationProvider.get_provider(provider_id)` - Get active provider
+- `IntegrationProvider.get_api_key(provider_id)` - Get API key for provider
+- `IntegrationProvider.get_providers_by_type(type)` - List providers by type
+
+---
+
+### AIModelConfig (NEW v1.4.0)
+
+Single Source of Truth for all AI models with pricing and credit configuration.
+
+```python
+class AIModelConfig(models.Model):
+    """Per final-model-schemas.md - Model definitions + pricing"""
+    model_name = CharField(max_length=100, unique=True)  # gpt-4o-mini, dall-e-3, runware:97@1
+    model_type = CharField(max_length=20)  # text, image
+    provider = CharField(max_length=50)  # Links to IntegrationProvider
+    display_name = CharField(max_length=200)
+    
+    is_default = BooleanField(default=False)  # One default per type
+    is_active = BooleanField(default=True)
+    
+    # Text Model Pricing (per 1K tokens)
+    cost_per_1k_input = DecimalField(max_digits=10, decimal_places=6, null=True)
+    cost_per_1k_output = DecimalField(max_digits=10, decimal_places=6, null=True)
+    tokens_per_credit = IntegerField(null=True)  # e.g., 1000, 10000
+    
+    # Image Model Pricing
+    credits_per_image = IntegerField(null=True)  # e.g., 1, 5, 15
+    quality_tier = CharField(max_length=20, null=True)  # basic, quality, premium
+    
+    # Model Limits
+    max_tokens = IntegerField(null=True)
+    context_window = IntegerField(null=True)
+    capabilities = JSONField(default=dict)  # vision, function_calling, etc.
+    
+    created_at = DateTimeField(auto_now_add=True)
+    updated_at = DateTimeField(auto_now=True)
+```
+
+**Credit Configuration Examples:**
+
+| Model | Type | tokens_per_credit | credits_per_image | quality_tier |
+|-------|------|-------------------|-------------------|--------------|
+| gpt-4o | text | 1000 | - | - |
+| gpt-4o-mini | text | 10000 | - | - |
+| runware:97@1 | image | - | 1 | basic |
+| dall-e-3 | image | - | 5 | quality |
+| google:4@2 | image | - | 15 | premium |
+
+**Helper Methods:**
+- `AIModelConfig.get_default_text_model()` - Get default text model
+- `AIModelConfig.get_default_image_model()` - Get default image model
+- `AIModelConfig.get_image_models_by_tier()` - List image models by quality tier
+
+---
+
+### SystemAISettings (NEW v1.4.0)
+
+System-wide AI defaults. Singleton (pk=1).
+
+```python
+class SystemAISettings(models.Model):
+    """Per final-model-schemas.md - Renamed from GlobalIntegrationSettings"""
+    # AI Parameters
+    temperature = FloatField(default=0.7)  # 0.0-2.0
+    max_tokens = IntegerField(default=8192)
+    
+    # Image Generation Settings
+    image_style = CharField(max_length=30, default='photorealistic')
+    image_quality = CharField(max_length=20, default='standard')  # standard, hd
+    max_images_per_article = IntegerField(default=4)  # 1-8
+    image_size = CharField(max_length=20, default='1024x1024')
+    
+    # Audit
+    updated_by = ForeignKey(User, null=True)
+    updated_at = DateTimeField(auto_now=True)
+```
+
+**Image Style Choices:**
+- `photorealistic` - Ultra realistic photography
+- `illustration` - Digital illustration
+- `3d_render` - Computer generated 3D
+- `minimal_flat` - Minimal / Flat Design
+- `artistic` - Artistic / Painterly
+- `cartoon` - Cartoon / Stylized
+
+**Image Size Choices:**
+- `1024x1024` - Square
+- `1792x1024` - Landscape
+- `1024x1792` - Portrait
+
+**Helper Methods:**
+- `SystemAISettings.get_instance()` - Get singleton instance
+- `SystemAISettings.get_effective_temperature(account)` - Get with account override
+- `SystemAISettings.get_effective_image_style(account)` - Get with account override
+
+---
+
+### AccountSettings (Per-Account Overrides)
+
+Generic key-value store for account-specific settings.
+
+```python
+class AccountSettings(AccountBaseModel):
+    """Per final-model-schemas.md - Account overrides"""
+    key = CharField(max_length=100)  # Setting key
+    value = JSONField(default=dict)  # Setting value
+    
+    created_at = DateTimeField(auto_now_add=True)
+    updated_at = DateTimeField(auto_now=True)
+```
+
+**AI-Related Keys** (override SystemAISettings defaults):
+
+| Key | Type | Example | Notes |
+|-----|------|---------|-------|
+| `ai.temperature` | float | 0.8 | Override system default |
+| `ai.max_tokens` | int | 8192 | Override system default |
+| `ai.image_style` | string | "illustration" | Override system default |
+| `ai.image_quality` | string | "hd" | Override system default |
+| `ai.max_images` | int | 6 | Override system default |
+| `ai.image_quality_tier` | string | "premium" | User's preferred tier |
+
+---
+
+### CreditCostConfig (Operation-Level Pricing)
+
+Fixed credit costs per operation type.
+
+```python
+class CreditCostConfig(models.Model):
+    """Per final-model-schemas.md - Operation pricing"""
+    operation_type = CharField(max_length=50, primary_key=True)  # Unique operation ID
+    display_name = CharField(max_length=100)
+    base_credits = IntegerField(default=1)  # Fixed credits per operation
+    is_active = BooleanField(default=True)
+    description = TextField(blank=True)
+```
+
+**Note:** `tokens_per_credit` moved to AIModelConfig as of v1.4.0.
+
+---
+
+## DEPRECATED Models
+
+### GlobalIntegrationSettings (DEPRECATED in v1.4.0)
+
+**Replaced by:** `IntegrationProvider` (API keys) + `AIModelConfig` (model configs) + `SystemAISettings` (defaults)
+
+```python
+# DEPRECATED - Do not use
+class GlobalIntegrationSettings(models.Model):
+    # API keys now in IntegrationProvider
+    openai_api_key = CharField(max_length=500)  # → IntegrationProvider.get_api_key('openai')
+    runware_api_key = CharField(max_length=500)  # → IntegrationProvider.get_api_key('runware')
+    
+    # Models now in AIModelConfig
+    openai_model = CharField(default='gpt-4o-mini')  # → AIModelConfig.is_default
+    
+    # Settings now in SystemAISettings
+    openai_temperature = FloatField(default=0.7)  # → SystemAISettings.temperature
+    openai_max_tokens = IntegerField(default=8192)  # → SystemAISettings.max_tokens
+```
+
+---

 ### GlobalAIPrompt