diff --git a/INTEGRATION-SETTINGS-WORKFLOW.md b/INTEGRATION-SETTINGS-WORKFLOW.md
new file mode 100644
index 00000000..167b6187
--- /dev/null
+++ b/INTEGRATION-SETTINGS-WORKFLOW.md
@@ -0,0 +1,223 @@
+# Integration Settings Workflow & Data Flow
+
+## Part 1: How Global Settings Load on Frontend
+
+### Admin Configures Global Settings
+**URL**: `https://api.igny8.com/admin/system/globalintegrationsettings/1/change/`
+
+**What's Stored**:
+- Platform-wide API keys (OpenAI, DALL-E, Runware)
+- Default model selections (gpt-4o-mini, dall-e-3, runware:97@1)
+- Default parameters (temperature: 0.7, max_tokens: 8192)
+- Default image settings (size, quality, style)
+
+**Who Can Access**: Only platform administrators
+
+### Normal User Opens Integration Page
+**URL**: `https://app.igny8.com/settings/integration`
+
+**What Happens**:
+
+1. **Frontend Request**:
+   - User browser requests: `GET /api/v1/system/settings/integrations/openai/`
+   - User browser requests: `GET /api/v1/system/settings/integrations/image_generation/`
+
+2. **Backend Processing** (`integration_views.py` - `get_settings()` method):
+   - Checks if user's account has custom overrides in `IntegrationSettings` table
+   - Gets global defaults from `GlobalIntegrationSettings` singleton
+   - Merges data with this priority:
+     - If account has overrides → use account settings
+     - If no overrides → use global defaults
+   - **NEVER returns API keys** (security)
+
+3. **Response to Frontend**:
+   ```
+   {
+     "id": "openai",
+     "enabled": true,
+     "model": "gpt-4o-mini",           // From global OR account override
+     "temperature": 0.7,                // From global OR account override
+     "max_tokens": 8192,                // From global OR account override
+     "using_global": true               // Flag: true if using defaults
+   }
+   ```
+
+4. **Frontend Display**:
+   - Shows current model selection
+   - Shows "Using platform defaults" badge if `using_global: true`
+   - Shows "Custom settings" badge if `using_global: false`
+   - User can change model, temperature, etc.
+   - **API key status is NOT shown** (user cannot see/change platform keys)
+
+---
+
+## Part 2: How User Changes Are Saved
+
+### User Changes Settings on Frontend
+
+1. **User Actions**:
+   - Opens settings modal
+   - Changes model from `gpt-4o-mini` to `gpt-4o`
+   - Changes temperature from `0.7` to `0.8`
+   - Clicks "Save"
+
+2. **Frontend Request**:
+   - Sends: `PUT /api/v1/system/settings/integrations/openai/`
+   - Body: `{"model": "gpt-4o", "temperature": 0.8, "max_tokens": 8192}`
+
+3. **Backend Processing** (`integration_views.py` - `save_settings()` method):
+   - **CRITICAL SECURITY**: Strips ANY API keys from request (apiKey, api_key, openai_api_key, etc.)
+   - Validates account exists
+   - Builds clean config with ONLY allowed overrides:
+     - For OpenAI: model, temperature, max_tokens
+     - For Image: service, model, image_quality, image_style, sizes
+   - Saves to `IntegrationSettings` table:
+     ```
+     account_id: 123
+     integration_type: "openai"
+     config: {"model": "gpt-4o", "temperature": 0.8, "max_tokens": 8192}
+     is_active: true
+     ```
+
+4. **Database Structure**:
+   - **GlobalIntegrationSettings** (1 row, pk=1):
+     - Contains: API keys + default settings
+     - Used by: ALL accounts for API keys
+   
+   - **IntegrationSettings** (multiple rows):
+     - Row per account per integration type
+     - Contains: ONLY overrides (no API keys)
+     - Example:
+       ```
+       id  | account_id | integration_type | config
+       100 | 123        | openai          | {"model": "gpt-4o", "temperature": 0.8}
+       101 | 456        | openai          | {"model": "gpt-4.1", "max_tokens": 4000}
+       102 | 123        | image_generation| {"service": "runware", "model": "runware:100@1"}
+       ```
+
+5. **Next Request from User**:
+   - Frontend requests: `GET /api/v1/system/settings/integrations/openai/`
+   - Backend finds IntegrationSettings row for account 123
+   - Returns: `{"model": "gpt-4o", "temperature": 0.8, "using_global": false}`
+   - User sees their custom settings
+
+---
+
+## Data Flow Architecture
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│  ADMIN SIDE                                                 │
+│  https://api.igny8.com/admin/                               │
+│                                                             │
+│  GlobalIntegrationSettings (pk=1)                           │
+│  ├── openai_api_key: "sk-xxx"        ← Platform-wide       │
+│  ├── openai_model: "gpt-4o-mini"     ← Default             │
+│  ├── openai_temperature: 0.7         ← Default             │
+│  ├── dalle_api_key: "sk-xxx"         ← Platform-wide       │
+│  ├── runware_api_key: "xxx"          ← Platform-wide       │
+│  └── image_quality: "standard"       ← Default             │
+└─────────────────────────────────────────────────────────────┘
+                            │
+                            │ Backend reads from
+                            ↓
+┌─────────────────────────────────────────────────────────────┐
+│  BACKEND API LAYER                                          │
+│  integration_views.py                                       │
+│                                                             │
+│  get_settings():                                            │
+│  1. Load GlobalIntegrationSettings (for defaults)           │
+│  2. Check IntegrationSettings (for account overrides)       │
+│  3. Merge: account overrides > global defaults              │
+│  4. Return to frontend (NO API keys)                        │
+│                                                             │
+│  save_settings():                                           │
+│  1. Receive request from frontend                           │
+│  2. Strip ALL API keys (security)                           │
+│  3. Save ONLY overrides to IntegrationSettings              │
+└─────────────────────────────────────────────────────────────┘
+                            │
+                            │ API sends data
+                            ↓
+┌─────────────────────────────────────────────────────────────┐
+│  FRONTEND - USER SIDE                                       │
+│  https://app.igny8.com/settings/integration                 │
+│                                                             │
+│  User sees:                                                 │
+│  ├── Model: gpt-4o-mini (dropdown)                          │
+│  ├── Temperature: 0.7 (slider)                              │
+│  ├── Status: ✓ Connected (test connection works)           │
+│  └── Badge: "Using platform defaults"                       │
+│                                                             │
+│  User CANNOT see:                                           │
+│  ✗ API keys (security)                                      │
+│  ✗ Platform configuration                                   │
+└─────────────────────────────────────────────────────────────┘
+                            │
+                            │ User changes settings
+                            ↓
+┌─────────────────────────────────────────────────────────────┐
+│  IntegrationSettings Table                                  │
+│  (Per-account overrides - NO API KEYS)                      │
+│                                                             │
+│  Account 123:                                               │
+│  ├── openai: {"model": "gpt-4o", "temperature": 0.8}       │
+│  └── image_generation: {"service": "runware"}               │
+│                                                             │
+│  Account 456:                                               │
+│  ├── openai: {"model": "gpt-4.1"}                           │
+│  └── image_generation: (no row = uses global defaults)      │
+└─────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## Important Security Rules
+
+1. **API Keys Flow**:
+   - Admin sets → GlobalIntegrationSettings
+   - Backend uses → For ALL accounts
+   - Frontend NEVER sees → Security
+   - Users NEVER save → Stripped by backend
+
+2. **Settings Flow**:
+   - Admin sets defaults → GlobalIntegrationSettings
+   - Users customize → IntegrationSettings (overrides only)
+   - Backend merges → Global defaults + account overrides
+   - Frontend displays → Merged result (no keys)
+
+3. **Free Plan Restriction**:
+   - Cannot create IntegrationSettings rows
+   - Must use global defaults only
+   - Enforced at frontend (UI disabled)
+   - TODO: Add backend validation
+
+---
+
+## Example Scenarios
+
+### Scenario 1: New User First Visit
+- User has NO IntegrationSettings row
+- Backend returns global defaults
+- `using_global: true`
+- User sees platform defaults
+- API operations use platform API key
+
+### Scenario 2: User Customizes Model
+- User changes model to "gpt-4o"
+- Frontend sends: `{"model": "gpt-4o"}`
+- Backend creates IntegrationSettings row
+- Next visit: `using_global: false`
+- API operations use platform API key + user's model choice
+
+### Scenario 3: User Resets to Default
+- Frontend sends: `{"model": "gpt-4o-mini"}` (same as global)
+- Backend still saves override row
+- Alternative: Delete row to truly use global
+- TODO: Add "Reset to defaults" button
+
+### Scenario 4: Admin Changes Global Default
+- Admin changes global model to "gpt-4.1"
+- Users WITH overrides: See their custom model
+- Users WITHOUT overrides: See new "gpt-4.1" default
+- All users: Use platform API key
diff --git a/backend/igny8_core/admin/site.py b/backend/igny8_core/admin/site.py
index f995d12d..fe3a383c 100644
--- a/backend/igny8_core/admin/site.py
+++ b/backend/igny8_core/admin/site.py
@@ -157,6 +157,14 @@ class Igny8AdminSite(UnfoldAdminSite):
                     ('igny8_core_auth', 'SeedKeyword'),
                 ],
             },
+            'Global Settings': {
+                'models': [
+                    ('system', 'GlobalIntegrationSettings'),
+                    ('system', 'GlobalAIPrompt'),
+                    ('system', 'GlobalAuthorProfile'),
+                    ('system', 'GlobalStrategy'),
+                ],
+            },
             'Plans and Billing': {
                 'models': [
                     ('igny8_core_auth', 'Plan'),
diff --git a/backend/igny8_core/ai/prompts.py b/backend/igny8_core/ai/prompts.py
index d95e65bf..8d021f7c 100644
--- a/backend/igny8_core/ai/prompts.py
+++ b/backend/igny8_core/ai/prompts.py
@@ -1,6 +1,6 @@
 """
 Prompt Registry - Centralized prompt management with override hierarchy
-Supports: task-level overrides → DB prompts → default fallbacks
+Supports: task-level overrides → DB prompts → GlobalAIPrompt (REQUIRED)
 """
 import logging
 from typing import Dict, Any, Optional
@@ -14,583 +14,11 @@ class PromptRegistry:
     Centralized prompt registry with hierarchical resolution:
     1. Task-level prompt_override (if exists)
     2. DB prompt for (account, function)
-    3. Default fallback from registry
+    3. GlobalAIPrompt (REQUIRED - no hardcoded fallbacks)
     """
     
-    # Default prompts stored in registry
-    DEFAULT_PROMPTS = {
-        'clustering': """You are a semantic strategist and SEO architecture engine. Your task is to analyze the provided keyword list and group them into meaningful, intent-driven topic clusters that reflect how real users search, think, and act online.
-
-Return a single JSON object with a "clusters" array. Each cluster must follow this structure:
-
-{
-  "name": "[Descriptive cluster name — natural, SEO-relevant, clearly expressing the topic]",
-  "description": "[1–2 concise sentences explaining what this cluster covers and why these keywords belong together]",
-  "keywords": ["keyword 1", "keyword 2", "keyword 3", "..."]
-}
-
-CLUSTERING STRATEGY:
-
-1. Keyword-first, structure-follows:
-   - Do NOT rely on assumed categories or existing content structures.
-   - Begin purely from the meaning, intent, and behavioral connection between keywords.
-
-2. Use multi-dimensional grouping logic:
-   - Group keywords by these behavioral dimensions:
-     • Search Intent → informational, commercial, transactional, navigational  
-     • Use-Case or Problem → what the user is trying to achieve or solve  
-     • Function or Feature → how something works or what it does  
-     • Persona or Audience → who the content or product serves  
-     • Context → location, time, season, platform, or device  
-   - Combine 2–3 dimensions naturally where they make sense.
-
-3. Model real search behavior:
-   - Favor clusters that form natural user journeys such as:
-     • Problem ➝ Solution
-     • General ➝ Specific
-     • Product ➝ Use-case
-     • Buyer ➝ Benefit
-     • Tool ➝ Function
-     • Task ➝ Method
-   - Each cluster should feel like a real topic hub users would explore in depth.
-
-4. Avoid superficial groupings:
-   - Do not cluster keywords just because they share words.
-   - Do not force-fit outliers or unrelated keywords.
-   - Exclude keywords that don't logically connect to any cluster.
-
-5. Quality rules:
-   - Each cluster should include between 3–10 strongly related keywords.
-   - Never duplicate a keyword across multiple clusters.
-   - Prioritize semantic strength, search intent, and usefulness for SEO-driven content structure.
-   - It's better to output fewer, high-quality clusters than many weak or shallow ones.
-
-INPUT FORMAT:
-{
-  "keywords": [IGNY8_KEYWORDS]
-}
-
-OUTPUT FORMAT:
-Return ONLY the final JSON object in this format:
-{
-  "clusters": [
-    {
-      "name": "...",
-      "description": "...",
-      "keywords": ["...", "...", "..."]
-    }
-  ]
-}
-
-Do not include any explanations, text, or commentary outside the JSON output.
-""",
-        
-        'ideas': """Generate SEO-optimized, high-quality content ideas and outlines for each keyword cluster.
-Input:
-Clusters: [IGNY8_CLUSTERS]
-Keywords: [IGNY8_CLUSTER_KEYWORDS]
-
-Output: JSON with "ideas" array.
-Each cluster → 1 cluster_hub + 2–4 supporting ideas.
-Each idea must include:
-title, description, content_type, content_structure, cluster_id, estimated_word_count (1500–2200), and covered_keywords.
-
-Outline Rules:
-
-Intro: 1 hook (30–40 words) + 2 intro paragraphs (50–60 words each).
-
-5–8 H2 sections, each with 2–3 H3s.
-
-Each H2 ≈ 250–300 words, mixed content (paragraphs, lists, tables, blockquotes).
-
-Vary section format and tone; no bullets or lists at start.
-
-Tables have columns; blockquotes = expert POV or data insight.
-
-Use depth, examples, and real context.
-
-Avoid repetitive structure.
-
-Tone: Professional editorial flow. No generic phrasing. Use varied sentence openings and realistic examples.
-
-Output JSON Example:
-
-{
-  "ideas": [
-    {
-      "title": "Best Organic Cotton Duvet Covers for All Seasons",
-      "description": {
-        "introduction": {
-          "hook": "Transform your sleep with organic cotton that blends comfort and sustainability.",
-          "paragraphs": [
-            {"format": "paragraph", "details": "Overview of organic cotton's rise in bedding industry."},
-            {"format": "paragraph", "details": "Why consumers prefer organic bedding over synthetic alternatives."}
-          ]
-        },
-        "H2": [
-          {
-            "heading": "Why Choose Organic Cotton for Bedding?",
-            "subsections": [
-              {"subheading": "Health and Skin Benefits", "format": "paragraph", "details": "Discuss hypoallergenic and chemical-free aspects."},
-              {"subheading": "Environmental Sustainability", "format": "list", "details": "Eco benefits like low water use, no pesticides."},
-              {"subheading": "Long-Term Cost Savings", "format": "table", "details": "Compare durability and pricing over time."}
-            ]
-          }
-        ]
-      },
-      "content_type": "post",
-      "content_structure": "review",
-      "cluster_id": 12,
-      "estimated_word_count": 1800,
-      "covered_keywords": "organic duvet covers, eco-friendly bedding, sustainable sheets"
-    }
-  ]
-}
-
-Valid content_type values: post, page, product, taxonomy
-
-Valid content_structure by type:
-- post: article, guide, comparison, review, listicle
-- page: landing_page, business_page, service_page, general, cluster_hub
-- product: product_page
-- taxonomy: category_archive, tag_archive, attribute_archive""",
-        
-        'content_generation': """You are an editorial content strategist. Your task is to generate a complete JSON response object based on the provided content idea, keyword cluster, keyword list, and metadata context.
-
-==================
-Generate a complete JSON response object matching this structure:
-==================
-
-{
-  "title": "[Article title using target keywords — full sentence case]",
-  "content": "[HTML content — full editorial structure with <p>, <h2>, <h3>, <ul>, <ol>, <table>]"
-}
-
-===========================
-CONTENT FLOW RULES
-===========================
-
-**INTRODUCTION:**
-- Start with 1 italicized hook (30–40 words)
-- Follow with 2 narrative paragraphs (each 50–60 words; 2–3 sentences max)
-- No headings allowed in intro
-
-**H2 SECTIONS (5–8 total):**
-Each section should be 250–300 words and follow this format:
-1. Two narrative paragraphs (80–120 words each, 2–3 sentences)
-2. One list or table (must come *after* a paragraph)
-3. Optional closing paragraph (40–60 words)
-4. Insert 2–3  subsections naturally after main paragraphs
-
-**Formatting Rules:**
-- Vary use of unordered lists, ordered lists, and tables across sections
-- Never begin any section or sub-section with a list or table
-
-===========================
-STYLE & QUALITY RULES
-===========================
-
-- **Keyword Usage:**
-  - Use keywords naturally in title, introduction, and headings
-  - Prioritize readability over keyword density
-
-- **Tone & style guidelines:**
-  - No robotic or passive voice
-  - Avoid generic intros like "In today's world…"
-  - Don't repeat heading in opening sentence
-  - Vary sentence structure and length
-
-===========================
-STAGE 3: METADATA CONTEXT (NEW)
-===========================
-
-**Content Structure:**
-[IGNY8_CONTENT_STRUCTURE]
-- If structure is "cluster_hub": Create comprehensive, authoritative content that serves as the main resource for this topic cluster. Include overview sections, key concepts, and links to related topics.
-- If structure is "article" or "guide": Create detailed, focused content that dives deep into the topic with actionable insights.
-- Other structures: Follow the appropriate format (listicle, comparison, review, landing_page, service_page, product_page, category_archive, tag_archive, attribute_archive).
-
-**Taxonomy Context:**
-[IGNY8_TAXONOMY]
-- Use taxonomy information to structure categories and tags appropriately.
-- Align content with taxonomy hierarchy and relationships.
-- Ensure content fits within the defined taxonomy structure.
-
-**Product/Service Attributes:**
-[IGNY8_ATTRIBUTES]
-- If attributes are provided (e.g., product specs, service modifiers), incorporate them naturally into the content.
-- For product content: Include specifications, features, dimensions, materials, etc. as relevant.
-- For service content: Include service tiers, pricing modifiers, availability, etc. as relevant.
-- Present attributes in a user-friendly format (tables, lists, or integrated into narrative).
-
-===========================
-INPUT VARIABLES
-===========================
-
-CONTENT IDEA DETAILS:
-[IGNY8_IDEA]
-
-KEYWORD CLUSTER:
-[IGNY8_CLUSTER]
-
-ASSOCIATED KEYWORDS:
-[IGNY8_KEYWORDS]
-
-===========================
-OUTPUT FORMAT
-===========================
-
-Return ONLY the final JSON object.  
-Do NOT include any comments, formatting, or explanations.""",
-
-        'site_structure_generation': """You are a senior UX architect and information designer. Use the business brief, objectives, style references, and existing site info to propose a complete multi-page marketing website structure.
-
-INPUT CONTEXT
-==============
-BUSINESS BRIEF:
-[IGNY8_BUSINESS_BRIEF]
-
-PRIMARY OBJECTIVES:
-[IGNY8_OBJECTIVES]
-
-STYLE & BRAND NOTES:
-[IGNY8_STYLE]
-
-SITE INFO / CURRENT STRUCTURE:
-[IGNY8_SITE_INFO]
-
-OUTPUT REQUIREMENTS
-====================
-Return ONE JSON object with the following keys:
-
-{
-  "site": {
-    "name": "...",
-    "primary_navigation": ["home", "services", "about", "contact"],
-    "secondary_navigation": ["blog", "faq"],
-    "hero_message": "High level value statement",
-    "tone": "voice + tone summary"
-  },
-  "pages": [
-    {
-      "slug": "home",
-      "title": "Home",
-      "type": "home | about | services | products | blog | contact | custom",
-      "status": "draft",
-      "objective": "Explain the core brand promise and primary CTA",
-      "primary_cta": "Book a strategy call",
-      "seo": {
-        "meta_title": "...",
-        "meta_description": "..."
-      },
-      "blocks": [
-        {
-          "type": "hero | features | services | stats | testimonials | faq | contact | custom",
-          "heading": "Section headline",
-          "subheading": "Support copy",
-          "layout": "full-width | two-column | cards | carousel",
-          "content": [
-            "Bullet or short paragraph describing what to render in this block"
-          ]
-        }
-      ]
-    }
-  ]
-}
-
-RULES
-=====
-- Include 5–8 pages covering the complete buyer journey (awareness → evaluation → conversion → trust).
-- Every page must have at least 3 blocks with concrete guidance (no placeholders like "Lorem ipsum").
-- Use consistent slug naming, all lowercase with hyphens.
-- Type must match the allowed enum and reflect page intent.
-- Ensure the navigation arrays align with the page list.
-- Focus on practical descriptions that an engineering team can hand off directly to the Site Builder.
-
-Return ONLY valid JSON. No commentary, explanations, or Markdown.
-""",
-        
-        'image_prompt_extraction': """Extract image prompts from the following article content.
-
-ARTICLE TITLE: {title}
-
-ARTICLE CONTENT:
-{content}
-
-Extract image prompts for:
-1. Featured Image: One main image that represents the article topic
-2. In-Article Images: Up to {max_images} images that would be useful within the article content
-
-Return a JSON object with this structure:
-{{
-    "featured_prompt": "Detailed description of the featured image",
-    "in_article_prompts": [
-        "Description of first in-article image",
-        "Description of second in-article image",
-        ...
-    ]
-}}
-
-Make sure each prompt is detailed enough for image generation, describing the visual elements, style, mood, and composition.""",
-        
-        'image_prompt_template': '{image_type} image for blog post titled "{post_title}": {image_prompt}',
-        
-        'negative_prompt': 'text, watermark, logo, overlay, title, caption, writing on walls, writing on objects, UI, infographic elements, post title',
-        
-        'optimize_content': """You are an expert content optimizer specializing in SEO, readability, and engagement.
-
-Your task is to optimize the provided content to improve its SEO score, readability, and engagement metrics.
-
-CURRENT CONTENT:
-Title: {CONTENT_TITLE}
-Word Count: {WORD_COUNT}
-Source: {SOURCE}
-Primary Keyword: {PRIMARY_KEYWORD}
-Internal Links: {INTERNAL_LINKS_COUNT}
-
-CURRENT META DATA:
-Meta Title: {META_TITLE}
-Meta Description: {META_DESCRIPTION}
-
-CURRENT SCORES:
-{CURRENT_SCORES}
-
-HTML CONTENT:
-{HTML_CONTENT}
-
-OPTIMIZATION REQUIREMENTS:
-
-1. SEO Optimization:
-   - Ensure meta title is 30-60 characters (if provided)
-   - Ensure meta description is 120-160 characters (if provided)
-   - Optimize primary keyword usage (natural, not keyword stuffing)
-   - Improve heading structure (H1, H2, H3 hierarchy)
-   - Add internal links where relevant (maintain existing links)
-
-2. Readability:
-   - Average sentence length: 15-20 words
-   - Use clear, concise language
-   - Break up long paragraphs
-   - Use bullet points and lists where appropriate
-   - Ensure proper paragraph structure
-
-3. Engagement:
-   - Add compelling headings
-   - Include relevant images placeholders (alt text)
-   - Use engaging language
-   - Create clear call-to-action sections
-   - Improve content flow and structure
-
-OUTPUT FORMAT:
-Return ONLY a JSON object in this format:
-{{
-  "html_content": "[Optimized HTML content]",
-  "meta_title": "[Optimized meta title, 30-60 chars]",
-  "meta_description": "[Optimized meta description, 120-160 chars]",
-  "optimization_notes": "[Brief notes on what was optimized]"
-}}
-
-Do not include any explanations, text, or commentary outside the JSON output.
-""",
-        
-        # Phase 8: Universal Content Types
-        'product_generation': """You are a product content specialist. Generate comprehensive product content that includes detailed descriptions, features, specifications, pricing, and benefits.
-
-INPUT:
-Product Name: [IGNY8_PRODUCT_NAME]
-Product Description: [IGNY8_PRODUCT_DESCRIPTION]
-Product Features: [IGNY8_PRODUCT_FEATURES]
-Target Audience: [IGNY8_TARGET_AUDIENCE]
-Primary Keyword: [IGNY8_PRIMARY_KEYWORD]
-
-OUTPUT FORMAT:
-Return ONLY a JSON object in this format:
-{
-  "title": "[Product name and key benefit]",
-  "meta_title": "[SEO-optimized meta title, 30-60 chars]",
-  "meta_description": "[Compelling meta description, 120-160 chars]",
-  "html_content": "[Complete HTML product page content]",
-  "word_count": [Integer word count],
-  "primary_keyword": "[Primary keyword]",
-  "secondary_keywords": ["keyword1", "keyword2", "keyword3"],
-  "tags": ["tag1", "tag2", "tag3"],
-  "categories": ["Category > Subcategory"],
-  "json_blocks": [
-    {
-      "type": "product_overview",
-      "heading": "Product Overview",
-      "content": "Detailed product description"
-    },
-    {
-      "type": "features",
-      "heading": "Key Features",
-      "items": ["Feature 1", "Feature 2", "Feature 3"]
-    },
-    {
-      "type": "specifications",
-      "heading": "Specifications",
-      "data": {"Spec 1": "Value 1", "Spec 2": "Value 2"}
-    },
-    {
-      "type": "pricing",
-      "heading": "Pricing",
-      "content": "Pricing information"
-    },
-    {
-      "type": "benefits",
-      "heading": "Benefits",
-      "items": ["Benefit 1", "Benefit 2", "Benefit 3"]
-    }
-  ],
-  "structure_data": {
-    "product_type": "[Product type]",
-    "price_range": "[Price range]",
-    "target_market": "[Target market]"
-  }
-}
-
-CONTENT REQUIREMENTS:
-- Include compelling product overview
-- List key features with benefits
-- Provide detailed specifications
-- Include pricing information (if available)
-- Highlight unique selling points
-- Use SEO-optimized headings
-- Include call-to-action sections
-- Ensure natural keyword usage
-""",
-        
-        'service_generation': """You are a service page content specialist. Generate comprehensive service page content that explains services, benefits, process, and pricing.
-
-INPUT:
-Service Name: [IGNY8_SERVICE_NAME]
-Service Description: [IGNY8_SERVICE_DESCRIPTION]
-Service Benefits: [IGNY8_SERVICE_BENEFITS]
-Target Audience: [IGNY8_TARGET_AUDIENCE]
-Primary Keyword: [IGNY8_PRIMARY_KEYWORD]
-
-OUTPUT FORMAT:
-Return ONLY a JSON object in this format:
-{
-  "title": "[Service name and value proposition]",
-  "meta_title": "[SEO-optimized meta title, 30-60 chars]",
-  "meta_description": "[Compelling meta description, 120-160 chars]",
-  "html_content": "[Complete HTML service page content]",
-  "word_count": [Integer word count],
-  "primary_keyword": "[Primary keyword]",
-  "secondary_keywords": ["keyword1", "keyword2", "keyword3"],
-  "tags": ["tag1", "tag2", "tag3"],
-  "categories": ["Category > Subcategory"],
-  "json_blocks": [
-    {
-      "type": "service_overview",
-      "heading": "Service Overview",
-      "content": "Detailed service description"
-    },
-    {
-      "type": "benefits",
-      "heading": "Benefits",
-      "items": ["Benefit 1", "Benefit 2", "Benefit 3"]
-    },
-    {
-      "type": "process",
-      "heading": "Our Process",
-      "steps": ["Step 1", "Step 2", "Step 3"]
-    },
-    {
-      "type": "pricing",
-      "heading": "Pricing",
-      "content": "Pricing information"
-    },
-    {
-      "type": "faq",
-      "heading": "Frequently Asked Questions",
-      "items": [{"question": "Q1", "answer": "A1"}]
-    }
-  ],
-  "structure_data": {
-    "service_type": "[Service type]",
-    "duration": "[Service duration]",
-    "target_market": "[Target market]"
-  }
-}
-
-CONTENT REQUIREMENTS:
-- Clear service overview and value proposition
-- Detailed benefits and outcomes
-- Step-by-step process explanation
-- Pricing information (if available)
-- FAQ section addressing common questions
-- Include testimonials or case studies (if applicable)
-- Use SEO-optimized headings
-- Include call-to-action sections
-""",
-        
-        'taxonomy_generation': """You are a taxonomy and categorization specialist. Generate comprehensive taxonomy page content that organizes and explains categories, tags, and hierarchical structures.
-
-INPUT:
-Taxonomy Name: [IGNY8_TAXONOMY_NAME]
-Taxonomy Description: [IGNY8_TAXONOMY_DESCRIPTION]
-Taxonomy Items: [IGNY8_TAXONOMY_ITEMS]
-Primary Keyword: [IGNY8_PRIMARY_KEYWORD]
-
-OUTPUT FORMAT:
-Return ONLY a JSON object in this format:
-{{
-  "title": "[Taxonomy name and purpose]",
-  "meta_title": "[SEO-optimized meta title, 30-60 chars]",
-  "meta_description": "[Compelling meta description, 120-160 chars]",
-  "html_content": "[Complete HTML taxonomy page content]",
-  "word_count": [Integer word count],
-  "primary_keyword": "[Primary keyword]",
-  "secondary_keywords": ["keyword1", "keyword2", "keyword3"],
-  "tags": ["tag1", "tag2", "tag3"],
-  "categories": ["Category > Subcategory"],
-  "json_blocks": [
-    {{
-      "type": "taxonomy_overview",
-      "heading": "Taxonomy Overview",
-      "content": "Detailed taxonomy description"
-    }},
-    {{
-      "type": "categories",
-      "heading": "Categories",
-      "items": [
-        {{
-          "name": "Category 1",
-          "description": "Category description",
-          "subcategories": ["Subcat 1", "Subcat 2"]
-        }}
-      ]
-    }},
-    {{
-      "type": "tags",
-      "heading": "Tags",
-      "items": ["Tag 1", "Tag 2", "Tag 3"]
-    }},
-    {{
-      "type": "hierarchy",
-      "heading": "Taxonomy Hierarchy",
-      "structure": {{"Level 1": {{"Level 2": ["Level 3"]}}}}
-    }}
-  ],
-  "structure_data": {{
-    "taxonomy_type": "[Taxonomy type]",
-    "item_count": [Integer],
-    "hierarchy_levels": [Integer]
-  }}
-}}
-
-CONTENT REQUIREMENTS:
-- Clear taxonomy overview and purpose
-- Organized category structure
-- Tag organization and relationships
-- Hierarchical structure visualization
-- SEO-optimized headings
-- Include navigation and organization benefits
-- Use clear, descriptive language
-""",
-    }
+    # Removed ALL hardcoded prompts - GlobalAIPrompt is now the ONLY source of default prompts
+    # To add/modify prompts, use Django admin: /admin/system/globalaiprompt/
     
     # Mapping from function names to prompt types
     FUNCTION_TO_PROMPT_TYPE = {
@@ -622,7 +50,7 @@ CONTENT REQUIREMENTS:
         Priority:
         1. task.prompt_override (if task provided and has override)
         2. DB prompt for (account, function)
-        3. Default fallback from registry
+        3. GlobalAIPrompt (REQUIRED - no hardcoded fallbacks)
         
         Args:
             function_name: AI function name (e.g., 'auto_cluster', 'generate_ideas')
@@ -642,7 +70,7 @@ CONTENT REQUIREMENTS:
         # Step 2: Get prompt type
         prompt_type = cls.FUNCTION_TO_PROMPT_TYPE.get(function_name, function_name)
         
-        # Step 3: Try DB prompt
+        # Step 3: Try DB prompt (account-specific)
         if account:
             try:
                 from igny8_core.modules.system.models import AIPrompt
@@ -651,18 +79,30 @@ CONTENT REQUIREMENTS:
                     prompt_type=prompt_type,
                     is_active=True
                 )
-                logger.info(f"Using DB prompt for {function_name} (account {account.id})")
+                logger.info(f"Using account-specific prompt for {function_name} (account {account.id})")
                 prompt = db_prompt.prompt_value
                 return cls._render_prompt(prompt, context or {})
             except Exception as e:
-                logger.debug(f"No DB prompt found for {function_name}: {e}")
+                logger.debug(f"No account-specific prompt found for {function_name}: {e}")
         
-        # Step 4: Use default fallback
-        prompt = cls.DEFAULT_PROMPTS.get(prompt_type, '')
-        if not prompt:
-            logger.warning(f"No default prompt found for {prompt_type}, using empty string")
-        
-        return cls._render_prompt(prompt, context or {})
+        # Step 4: Try GlobalAIPrompt (platform-wide default) - REQUIRED
+        try:
+            from igny8_core.modules.system.global_settings_models import GlobalAIPrompt
+            global_prompt = GlobalAIPrompt.objects.get(
+                prompt_type=prompt_type,
+                is_active=True
+            )
+            logger.info(f"Using global default prompt for {function_name} from GlobalAIPrompt")
+            prompt = global_prompt.prompt_value
+            return cls._render_prompt(prompt, context or {})
+        except Exception as e:
+            error_msg = (
+                f"ERROR: Global prompt '{prompt_type}' not found for function '{function_name}'. "
+                f"Please configure it in Django admin at: /admin/system/globalaiprompt/. "
+                f"Error: {e}"
+            )
+            logger.error(error_msg)
+            raise ValueError(error_msg)
     
     @classmethod
     def _render_prompt(cls, prompt_template: str, context: Dict[str, Any]) -> str:
@@ -728,8 +168,17 @@ CONTENT REQUIREMENTS:
             except Exception:
                 pass
         
-        # Use default
-        return cls.DEFAULT_PROMPTS.get(prompt_type, '')
+        # Try GlobalAIPrompt
+        try:
+            from igny8_core.modules.system.global_settings_models import GlobalAIPrompt
+            global_prompt = GlobalAIPrompt.objects.get(
+                prompt_type=prompt_type,
+                is_active=True
+            )
+            return global_prompt.prompt_value
+        except Exception:
+            # Fallback for image_prompt_template
+            return '{image_type} image for blog post titled "{post_title}": {image_prompt}'
     
     @classmethod
     def get_negative_prompt(cls, account: Optional[Any] = None) -> str:
@@ -752,8 +201,17 @@ CONTENT REQUIREMENTS:
             except Exception:
                 pass
         
-        # Use default
-        return cls.DEFAULT_PROMPTS.get(prompt_type, '')
+        # Try GlobalAIPrompt
+        try:
+            from igny8_core.modules.system.global_settings_models import GlobalAIPrompt
+            global_prompt = GlobalAIPrompt.objects.get(
+                prompt_type=prompt_type,
+                is_active=True
+            )
+            return global_prompt.prompt_value
+        except Exception:
+            # Fallback for negative_prompt
+            return 'text, watermark, logo, overlay, title, caption, writing on walls, writing on objects, UI, infographic elements, post title'
 
 
 # Convenience function for backward compatibility
diff --git a/backend/igny8_core/business/automation/migrations/0005_add_default_image_service.py b/backend/igny8_core/business/automation/migrations/0005_add_default_image_service.py
new file mode 100644
index 00000000..5470dc85
--- /dev/null
+++ b/backend/igny8_core/business/automation/migrations/0005_add_default_image_service.py
@@ -0,0 +1,18 @@
+# Generated by Django 5.2.9 on 2025-12-20 15:13
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('automation', '0004_add_pause_resume_cancel_fields'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='automationconfig',
+            name='stage_1_batch_size',
+            field=models.IntegerField(default=50, help_text='Keywords per batch'),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/admin.py b/backend/igny8_core/modules/system/admin.py
index 8e598d24..aeb9c3e1 100644
--- a/backend/igny8_core/modules/system/admin.py
+++ b/backend/igny8_core/modules/system/admin.py
@@ -343,18 +343,22 @@ class GlobalIntegrationSettingsAdmin(Igny8ModelAdmin):
             "fields": ("openai_api_key", "openai_model", "openai_temperature", "openai_max_tokens"),
             "description": "Global OpenAI configuration used by all accounts (unless overridden)"
         }),
-        ("DALL-E Settings", {
-            "fields": ("dalle_api_key", "dalle_model", "dalle_size", "dalle_quality", "dalle_style"),
-            "description": "Global DALL-E image generation configuration"
+        ("Image Generation - Default Service", {
+            "fields": ("default_image_service",),
+            "description": "Choose which image generation service is used by default for all accounts"
         }),
-        ("Anthropic Settings", {
-            "fields": ("anthropic_api_key", "anthropic_model"),
-            "description": "Global Anthropic Claude configuration"
+        ("Image Generation - DALL-E", {
+            "fields": ("dalle_api_key", "dalle_model", "dalle_size"),
+            "description": "Global DALL-E (OpenAI) image generation configuration"
         }),
-        ("Runware Settings", {
-            "fields": ("runware_api_key",),
+        ("Image Generation - Runware", {
+            "fields": ("runware_api_key", "runware_model"),
             "description": "Global Runware image generation configuration"
         }),
+        ("Universal Image Settings", {
+            "fields": ("image_quality", "image_style", "max_in_article_images", "desktop_image_size", "mobile_image_size"),
+            "description": "Image quality, style, and sizing settings that apply to ALL providers (DALL-E, Runware, etc.)"
+        }),
         ("Status", {
             "fields": ("is_active", "last_updated", "updated_by")
         }),
diff --git a/backend/igny8_core/modules/system/global_settings_models.py b/backend/igny8_core/modules/system/global_settings_models.py
index 8b2b43ae..f36bc4f3 100644
--- a/backend/igny8_core/modules/system/global_settings_models.py
+++ b/backend/igny8_core/modules/system/global_settings_models.py
@@ -20,6 +20,61 @@ class GlobalIntegrationSettings(models.Model):
     - Starter/Growth/Scale: Can override model, temperature, tokens, etc.
     """
     
+    OPENAI_MODEL_CHOICES = [
+        ('gpt-4.1', 'GPT-4.1 - $2.00 / $8.00 per 1M tokens'),
+        ('gpt-4o-mini', 'GPT-4o mini - $0.15 / $0.60 per 1M tokens'),
+        ('gpt-4o', 'GPT-4o - $2.50 / $10.00 per 1M tokens'),
+        ('gpt-4-turbo-preview', 'GPT-4 Turbo Preview - $10.00 / $30.00 per 1M tokens'),
+        ('gpt-5.1', 'GPT-5.1 - $1.25 / $10.00 per 1M tokens (16K)'),
+        ('gpt-5.2', 'GPT-5.2 - $1.75 / $14.00 per 1M tokens (16K)'),
+    ]
+    
+    DALLE_MODEL_CHOICES = [
+        ('dall-e-3', 'DALL·E 3 - $0.040 per image'),
+        ('dall-e-2', 'DALL·E 2 - $0.020 per image'),
+    ]
+    
+    DALLE_SIZE_CHOICES = [
+        ('1024x1024', '1024x1024 (Square)'),
+        ('1792x1024', '1792x1024 (Landscape)'),
+        ('1024x1792', '1024x1792 (Portrait)'),
+        ('512x512', '512x512 (Small Square)'),
+    ]
+    
+    DALLE_QUALITY_CHOICES = [
+        ('standard', 'Standard'),
+        ('hd', 'HD'),
+    ]
+    
+    DALLE_STYLE_CHOICES = [
+        ('vivid', 'Vivid'),
+        ('natural', 'Natural'),
+    ]
+    
+    RUNWARE_MODEL_CHOICES = [
+        ('runware:97@1', 'Runware 97@1 - Versatile Model'),
+        ('runware:100@1', 'Runware 100@1 - High Quality'),
+        ('runware:101@1', 'Runware 101@1 - Fast Generation'),
+    ]
+    
+    IMAGE_QUALITY_CHOICES = [
+        ('standard', 'Standard'),
+        ('hd', 'HD'),
+    ]
+    
+    IMAGE_STYLE_CHOICES = [
+        ('vivid', 'Vivid'),
+        ('natural', 'Natural'),
+        ('realistic', 'Realistic'),
+        ('artistic', 'Artistic'),
+        ('cartoon', 'Cartoon'),
+    ]
+    
+    IMAGE_SERVICE_CHOICES = [
+        ('openai', 'OpenAI DALL-E'),
+        ('runware', 'Runware'),
+    ]
+    
     # OpenAI Settings (for text generation)
     openai_api_key = models.CharField(
         max_length=500,
@@ -28,7 +83,8 @@ class GlobalIntegrationSettings(models.Model):
     )
     openai_model = models.CharField(
         max_length=100,
-        default='gpt-4-turbo-preview',
+        default='gpt-4o-mini',
+        choices=OPENAI_MODEL_CHOICES,
         help_text="Default text generation model (accounts can override if plan allows)"
     )
     openai_temperature = models.FloatField(
@@ -40,7 +96,7 @@ class GlobalIntegrationSettings(models.Model):
         help_text="Default max tokens for responses (accounts can override if plan allows)"
     )
     
-    # DALL-E Settings (for image generation)
+    # Image Generation Settings (OpenAI/DALL-E)
     dalle_api_key = models.CharField(
         max_length=500,
         blank=True,
@@ -49,44 +105,64 @@ class GlobalIntegrationSettings(models.Model):
     dalle_model = models.CharField(
         max_length=100,
         default='dall-e-3',
+        choices=DALLE_MODEL_CHOICES,
         help_text="Default DALL-E model (accounts can override if plan allows)"
     )
     dalle_size = models.CharField(
         max_length=20,
         default='1024x1024',
+        choices=DALLE_SIZE_CHOICES,
         help_text="Default image size (accounts can override if plan allows)"
     )
-    dalle_quality = models.CharField(
-        max_length=20,
-        default='standard',
-        choices=[('standard', 'Standard'), ('hd', 'HD')],
-        help_text="Default image quality (accounts can override if plan allows)"
-    )
-    dalle_style = models.CharField(
-        max_length=20,
-        default='vivid',
-        choices=[('vivid', 'Vivid'), ('natural', 'Natural')],
-        help_text="Default image style (accounts can override if plan allows)"
-    )
     
-    # Anthropic Settings (for Claude)
-    anthropic_api_key = models.CharField(
-        max_length=500,
-        blank=True,
-        help_text="Platform Anthropic API key - used by ALL accounts"
-    )
-    anthropic_model = models.CharField(
-        max_length=100,
-        default='claude-3-sonnet-20240229',
-        help_text="Default Anthropic model (accounts can override if plan allows)"
-    )
-    
-    # Runware Settings (alternative image generation)
+    # Image Generation Settings (Runware)
     runware_api_key = models.CharField(
         max_length=500,
         blank=True,
         help_text="Platform Runware API key - used by ALL accounts"
     )
+    runware_model = models.CharField(
+        max_length=100,
+        default='runware:97@1',
+        choices=RUNWARE_MODEL_CHOICES,
+        help_text="Default Runware model (accounts can override if plan allows)"
+    )
+    
+    # Default Image Generation Service
+    default_image_service = models.CharField(
+        max_length=20,
+        default='openai',
+        choices=IMAGE_SERVICE_CHOICES,
+        help_text="Default image generation service for all accounts (openai=DALL-E, runware=Runware)"
+    )
+    
+    # Universal Image Generation Settings (applies to ALL providers)
+    image_quality = models.CharField(
+        max_length=20,
+        default='standard',
+        choices=IMAGE_QUALITY_CHOICES,
+        help_text="Default image quality for all providers (accounts can override if plan allows)"
+    )
+    image_style = models.CharField(
+        max_length=20,
+        default='realistic',
+        choices=IMAGE_STYLE_CHOICES,
+        help_text="Default image style for all providers (accounts can override if plan allows)"
+    )
+    max_in_article_images = models.IntegerField(
+        default=2,
+        help_text="Default maximum images to generate per article (1-5, accounts can override if plan allows)"
+    )
+    desktop_image_size = models.CharField(
+        max_length=20,
+        default='1024x1024',
+        help_text="Default desktop image size (accounts can override if plan allows)"
+    )
+    mobile_image_size = models.CharField(
+        max_length=20,
+        default='512x512',
+        help_text="Default mobile image size (accounts can override if plan allows)"
+    )
     
     # Metadata
     is_active = models.BooleanField(default=True)
@@ -151,7 +227,8 @@ class GlobalAIPrompt(models.Model):
     description = models.TextField(blank=True, help_text="Description of what this prompt does")
     variables = models.JSONField(
         default=list,
-        help_text="List of variables used in the prompt (e.g., {keyword}, {industry})"
+        blank=True,
+        help_text="Optional: List of variables used in the prompt (e.g., {keyword}, {industry})"
     )
     is_active = models.BooleanField(default=True, db_index=True)
     version = models.IntegerField(default=1, help_text="Prompt version for tracking changes")
diff --git a/backend/igny8_core/modules/system/integration_views.py b/backend/igny8_core/modules/system/integration_views.py
index 305e41ff..002a828a 100644
--- a/backend/igny8_core/modules/system/integration_views.py
+++ b/backend/igny8_core/modules/system/integration_views.py
@@ -94,14 +94,12 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
             permission_classes=[IsAuthenticatedAndActive, HasTenantAccess, IsAdminOrOwner])
     def test_connection(self, request, pk=None):
         """
-        Test API connection for OpenAI or Runware
-        Supports two modes:
-        - with_response=false: Simple connection test (GET /v1/models)
-        - with_response=true: Full response test with ping message
+        Test API connection using platform API keys.
+        Tests OpenAI or Runware with current model selection.
         """
         integration_type = pk  # 'openai', 'runware'
         
-        logger.info(f"[test_connection] Called for integration_type={integration_type}, user={getattr(request, 'user', None)}, account={getattr(request, 'account', None)}")
+        logger.info(f"[test_connection] Called for integration_type={integration_type}")
         
         if not integration_type:
             return error_response(
@@ -110,70 +108,43 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
                 request=request
             )
         
-        # Get API key and config from request or saved settings
-        config = request.data.get('config', {}) if isinstance(request.data.get('config'), dict) else {}
-        api_key = request.data.get('apiKey') or config.get('apiKey')
-        
-        # Merge request.data with config if config is a dict
-        if not isinstance(config, dict):
-            config = {}
-        
-        if not api_key:
-            # Try to get from saved settings
-            account = getattr(request, 'account', None)
-            logger.info(f"[test_connection] Account from request: {account.id if account else None}")
-            # Fallback to user's account
-            if not account:
-                user = getattr(request, 'user', None)
-                if user and hasattr(user, 'is_authenticated') and user.is_authenticated:
-                    account = getattr(user, 'account', None)
-            # Fallback to default account
-            if not account:
-                from igny8_core.auth.models import Account
-                try:
-                    account = Account.objects.first()
-                except Exception:
-                    pass
-            
-            if account:
-                try:
-                    from .models import IntegrationSettings
-                    logger.info(f"[test_connection] Looking for saved settings for account {account.id}")
-                    saved_settings = IntegrationSettings.objects.get(
-                        integration_type=integration_type,
-                        account=account
-                    )
-                    api_key = saved_settings.config.get('apiKey')
-                    logger.info(f"[test_connection] Found saved settings, has_apiKey={bool(api_key)}")
-                except IntegrationSettings.DoesNotExist:
-                    logger.warning(f"[test_connection] No saved settings found for {integration_type} and account {account.id}")
-                    pass
-        
-        if not api_key:
-            logger.error(f"[test_connection] No API key found in request or saved settings")
-            return error_response(
-                error='API key is required',
-                status_code=status.HTTP_400_BAD_REQUEST,
-                request=request
-            )
-        
-        logger.info(f"[test_connection] Testing {integration_type} connection with API key (length={len(api_key) if api_key else 0})")
         try:
+            from igny8_core.modules.system.global_settings_models import GlobalIntegrationSettings
+            
+            # Get platform API keys
+            global_settings = GlobalIntegrationSettings.get_instance()
+            
+            # Get config from request (model selection)
+            config = request.data.get('config', {}) if isinstance(request.data.get('config'), dict) else {}
+            
             if integration_type == 'openai':
+                api_key = global_settings.openai_api_key
+                if not api_key:
+                    return error_response(
+                        error='Platform OpenAI API key not configured. Please contact administrator.',
+                        status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+                        request=request
+                    )
                 return self._test_openai(api_key, config, request)
+                
             elif integration_type == 'runware':
+                api_key = global_settings.runware_api_key
+                if not api_key:
+                    return error_response(
+                        error='Platform Runware API key not configured. Please contact administrator.',
+                        status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+                        request=request
+                    )
                 return self._test_runware(api_key, request)
+                
             else:
                 return error_response(
-                    error=f'Validation not supported for {integration_type}',
+                    error=f'Testing not supported for {integration_type}',
                     status_code=status.HTTP_400_BAD_REQUEST,
                     request=request
                 )
         except Exception as e:
             logger.error(f"Error testing {integration_type} connection: {str(e)}", exc_info=True)
-            import traceback
-            error_trace = traceback.format_exc()
-            logger.error(f"Full traceback: {error_trace}")
             return error_response(
                 error=str(e),
                 status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
@@ -662,8 +633,13 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
         return self.save_settings(request, integration_type)
     
     def save_settings(self, request, pk=None):
-        """Save integration settings"""
-        integration_type = pk  # 'openai', 'runware', 'gsc'
+        """
+        Save integration settings (account overrides only).
+        - Saves model/parameter overrides to IntegrationSettings
+        - NEVER saves API keys (those are platform-wide)
+        - Free plan: Should be blocked at frontend level
+        """
+        integration_type = pk
         
         logger.info(f"[save_settings] Called for integration_type={integration_type}, user={getattr(request, 'user', None)}, account={getattr(request, 'account', None)}")
         
@@ -678,12 +654,19 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
         config = dict(request.data) if hasattr(request.data, 'dict') else (request.data if isinstance(request.data, dict) else {})
         logger.info(f"[save_settings] Config keys: {list(config.keys()) if isinstance(config, dict) else 'Not a dict'}")
         
+        # Remove any API keys from config (security - they shouldn't be sent but just in case)
+        config.pop('apiKey', None)
+        config.pop('api_key', None)
+        config.pop('openai_api_key', None)
+        config.pop('dalle_api_key', None)
+        config.pop('runware_api_key', None)
+        config.pop('anthropic_api_key', None)
+        
         try:
-            # Get account - try multiple methods
+            # Get account
             account = getattr(request, 'account', None)
             logger.info(f"[save_settings] Account from request: {account.id if account else None}")
             
-            # Fallback 1: Get from authenticated user's account
             if not account:
                 user = getattr(request, 'user', None)
                 if user and hasattr(user, 'is_authenticated') and user.is_authenticated:
@@ -693,93 +676,81 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
                         logger.warning(f"Error getting account from user: {e}")
                         account = None
             
-            # Fallback 2: If still no account, get default account (for development)
             if not account:
-                from igny8_core.auth.models import Account
-                try:
-                    # Get the first account as fallback (development only)
-                    account = Account.objects.first()
-                except Exception as e:
-                    logger.warning(f"Error getting default account: {e}")
-                    account = None
-            
-            if not account:
-                logger.error(f"[save_settings] No account found after all fallbacks")
+                logger.error(f"[save_settings] No account found")
                 return error_response(
                     error='Account not found. Please ensure you are logged in.',
                     status_code=status.HTTP_400_BAD_REQUEST,
                     request=request
                 )
             
-            logger.info(f"[save_settings] Using account: {account.id} ({account.name}, slug={account.slug}, status={account.status})")
+            logger.info(f"[save_settings] Using account: {account.id} ({account.name})")
             
-            # Store integration settings in a simple model or settings table
-            # For now, we'll use a simple approach - store in IntegrationSettings model
-            # or use Django settings/database
+            # TODO: Check if Free plan - they shouldn't be able to save overrides
+            # This should be blocked at frontend level, but add backend check too
             
-            # Import IntegrationSettings model
             from .models import IntegrationSettings
                 
-            # For image_generation, ensure provider is set correctly
-            if integration_type == 'image_generation':
+            # Build clean config with only allowed overrides
+            clean_config = {}
+            
+            if integration_type == 'openai':
+                # Only allow model, temperature, max_tokens overrides
+                if 'model' in config:
+                    clean_config['model'] = config['model']
+                if 'temperature' in config:
+                    clean_config['temperature'] = config['temperature']
+                if 'max_tokens' in config:
+                    clean_config['max_tokens'] = config['max_tokens']
+                    
+            elif integration_type == 'image_generation':
                 # Map service to provider if service is provided
-                if 'service' in config and 'provider' not in config:
-                    config['provider'] = config['service']
-                # Ensure provider is set
-                if 'provider' not in config:
-                    config['provider'] = config.get('service', 'openai')
-                # Set model based on provider
-                if config.get('provider') == 'openai' and 'model' not in config:
-                    config['model'] = config.get('imageModel', 'dall-e-3')
-                elif config.get('provider') == 'runware' and 'model' not in config:
-                    config['model'] = config.get('runwareModel', 'runware:97@1')
-                # Ensure all image settings have defaults (except max_in_article_images which must be explicitly set)
-                config.setdefault('image_type', 'realistic')
-                config.setdefault('image_format', 'webp')
-                config.setdefault('desktop_enabled', True)
-                config.setdefault('mobile_enabled', True)
-                
-                # Set default image sizes based on provider/model
-                provider = config.get('provider', 'openai')
-                model = config.get('model', 'dall-e-3')
-                
-                if not config.get('featured_image_size'):
-                    if provider == 'runware':
-                        config['featured_image_size'] = '1280x832'
-                    else:  # openai
-                        config['featured_image_size'] = '1024x1024'
-                
-                if not config.get('desktop_image_size'):
-                    config['desktop_image_size'] = '1024x1024'
+                if 'service' in config:
+                    clean_config['service'] = config['service']
+                    clean_config['provider'] = config['service']
+                if 'provider' in config:
+                    clean_config['provider'] = config['provider']
+                    clean_config['service'] = config['provider']
+                    
+                # Model selection (service-specific)
+                if 'model' in config:
+                    clean_config['model'] = config['model']
+                if 'imageModel' in config:
+                    clean_config['imageModel'] = config['imageModel']
+                    clean_config['model'] = config['imageModel']  # Also store in 'model' for consistency
+                if 'runwareModel' in config:
+                    clean_config['runwareModel'] = config['runwareModel']
+                    
+                # Universal image settings (applies to all providers)
+                for key in ['image_type', 'image_quality', 'image_style', 'max_in_article_images', 'image_format', 
+                           'desktop_enabled', 'mobile_enabled', 'featured_image_size', 'desktop_image_size']:
+                    if key in config:
+                        clean_config[key] = config[key]
             
             # Get or create integration settings
-            logger.info(f"[save_settings] Attempting get_or_create for {integration_type} with account {account.id}")
+            logger.info(f"[save_settings] Saving clean config: {clean_config}")
             integration_settings, created = IntegrationSettings.objects.get_or_create(
                 integration_type=integration_type,
                 account=account,
-                defaults={'config': config, 'is_active': config.get('enabled', False)}
+                defaults={'config': clean_config, 'is_active': True}
             )
-            logger.info(f"[save_settings] get_or_create result: created={created}, id={integration_settings.id}")
+            logger.info(f"[save_settings] Result: created={created}, id={integration_settings.id}")
             
             if not created:
-                logger.info(f"[save_settings] Updating existing settings (id={integration_settings.id})")
-                integration_settings.config = config
-                integration_settings.is_active = config.get('enabled', False)
+                integration_settings.config = clean_config
+                integration_settings.is_active = True
                 integration_settings.save()
-                logger.info(f"[save_settings] Settings updated successfully")
+                logger.info(f"[save_settings] Updated existing settings")
             
-            logger.info(f"[save_settings] Successfully saved settings for {integration_type}")
+            logger.info(f"[save_settings] Successfully saved overrides for {integration_type}")
             return success_response(
-                data={'config': config},
+                data={'config': clean_config},
                 message=f'{integration_type.upper()} settings saved successfully',
                 request=request
             )
                 
         except Exception as e:
             logger.error(f"Error saving integration settings for {integration_type}: {str(e)}", exc_info=True)
-            import traceback
-            error_trace = traceback.format_exc()
-            logger.error(f"Full traceback: {error_trace}")
             return error_response(
                 error=f'Failed to save settings: {str(e)}',
                 status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
@@ -787,7 +758,13 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
             )
     
     def get_settings(self, request, pk=None):
-        """Get integration settings - defaults to AWS-admin settings if account doesn't have its own"""
+        """
+        Get integration settings for frontend.
+        Returns:
+        - Global defaults (model, temperature, etc.)
+        - Account overrides if they exist  
+        - NO API keys (platform-wide only)
+        """
         integration_type = pk
         
         if not integration_type:
@@ -798,10 +775,9 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
             )
         
         try:
-            # Get account - try multiple methods (same as save_settings)
+            # Get account
             account = getattr(request, 'account', None)
             
-            # Fallback 1: Get from authenticated user's account
             if not account:
                 user = getattr(request, 'user', None)
                 if user and hasattr(user, 'is_authenticated') and user.is_authenticated:
@@ -812,31 +788,116 @@ class IntegrationSettingsViewSet(viewsets.ViewSet):
                         account = None
             
             from .models import IntegrationSettings
+            from igny8_core.modules.system.global_settings_models import GlobalIntegrationSettings
             
-            # Get account-specific settings
-            if account:
-                try:
-                    integration_settings = IntegrationSettings.objects.get(
-                        integration_type=integration_type,
-                        account=account
-                    )
-                    response_data = {
-                        'id': integration_settings.integration_type,
-                        'enabled': integration_settings.is_active,
-                        **integration_settings.config
-                    }
-                    return success_response(
-                        data=response_data,
-                        request=request
-                    )
-                except IntegrationSettings.DoesNotExist:
-                    pass
-                except Exception as e:
-                    logger.error(f"Error getting account-specific settings: {e}", exc_info=True)
+            # Get global defaults
+            global_settings = GlobalIntegrationSettings.get_instance()
+            
+            # Build response with global defaults
+            if integration_type == 'openai':
+                response_data = {
+                    'id': 'openai',
+                    'enabled': True,  # Always enabled (platform-wide)
+                    'model': global_settings.openai_model,
+                    'temperature': global_settings.openai_temperature,
+                    'max_tokens': global_settings.openai_max_tokens,
+                    'using_global': True,  # Flag to show it's using global
+                }
+                
+                # Check for account overrides
+                if account:
+                    try:
+                        integration_settings = IntegrationSettings.objects.get(
+                            integration_type=integration_type,
+                            account=account,
+                            is_active=True
+                        )
+                        config = integration_settings.config or {}
+                        if config.get('model'):
+                            response_data['model'] = config['model']
+                            response_data['using_global'] = False
+                        if config.get('temperature') is not None:
+                            response_data['temperature'] = config['temperature']
+                        if config.get('max_tokens'):
+                            response_data['max_tokens'] = config['max_tokens']
+                    except IntegrationSettings.DoesNotExist:
+                        pass
+                
+            elif integration_type == 'runware':
+                response_data = {
+                    'id': 'runware',
+                    'enabled': True,  # Always enabled (platform-wide)
+                    'using_global': True,
+                }
+                
+            elif integration_type == 'image_generation':
+                # Get default service and model based on global settings
+                default_service = global_settings.default_image_service
+                default_model = global_settings.dalle_model if default_service == 'openai' else global_settings.runware_model
+                
+                response_data = {
+                    'id': 'image_generation',
+                    'enabled': True,
+                    'service': default_service,  # From global settings
+                    'provider': default_service,  # Alias for service
+                    'model': default_model,  # Service-specific default model
+                    'imageModel': global_settings.dalle_model,  # OpenAI model
+                    'runwareModel': global_settings.runware_model,  # Runware model
+                    'image_type': global_settings.image_style,  # Use image_style as default
+                    'image_quality': global_settings.image_quality,  # Universal quality
+                    'image_style': global_settings.image_style,  # Universal style
+                    'max_in_article_images': global_settings.max_in_article_images,
+                    'image_format': 'webp',
+                    'desktop_enabled': True,
+                    'mobile_enabled': True,
+                    'featured_image_size': global_settings.dalle_size,
+                    'desktop_image_size': global_settings.desktop_image_size,
+                    'mobile_image_size': global_settings.mobile_image_size,
+                    'using_global': True,
+                }
+                
+                # Check for account overrides
+                if account:
+                    try:
+                        integration_settings = IntegrationSettings.objects.get(
+                            integration_type=integration_type,
+                            account=account,
+                            is_active=True
+                        )
+                        config = integration_settings.config or {}
+                        # Override with account settings
+                        if config:
+                            response_data['using_global'] = False
+                            # Service/provider
+                            if 'service' in config:
+                                response_data['service'] = config['service']
+                                response_data['provider'] = config['service']
+                            if 'provider' in config:
+                                response_data['provider'] = config['provider']
+                                response_data['service'] = config['provider']
+                            # Models
+                            if 'model' in config:
+                                response_data['model'] = config['model']
+                            if 'imageModel' in config:
+                                response_data['imageModel'] = config['imageModel']
+                            if 'runwareModel' in config:
+                                response_data['runwareModel'] = config['runwareModel']
+                            # Universal image settings
+                            for key in ['image_type', 'image_quality', 'image_style', 'max_in_article_images', 'image_format', 
+                                       'desktop_enabled', 'mobile_enabled', 'featured_image_size', 'desktop_image_size']:
+                                if key in config:
+                                    response_data[key] = config[key]
+                    except IntegrationSettings.DoesNotExist:
+                        pass
+            else:
+                # Other integration types - return empty
+                response_data = {
+                    'id': integration_type,
+                    'enabled': False,
+                }
             
-            # Return empty config if no settings found
             return success_response(
-                data={},
+                data=response_data,
                 request=request
             )
         except Exception as e:
diff --git a/backend/igny8_core/modules/system/migrations/0004_fix_global_settings_remove_override.py b/backend/igny8_core/modules/system/migrations/0004_fix_global_settings_remove_override.py
index 8d80a79d..ec6caa2b 100644
--- a/backend/igny8_core/modules/system/migrations/0004_fix_global_settings_remove_override.py
+++ b/backend/igny8_core/modules/system/migrations/0004_fix_global_settings_remove_override.py
@@ -6,7 +6,7 @@ from django.db import migrations, models
 class Migration(migrations.Migration):
 
     dependencies = [
-        ('system', '0003_fix_global_settings_architecture'),
+        ('system', '0002_add_global_settings_models'),
     ]
 
     operations = [
diff --git a/backend/igny8_core/modules/system/migrations/0005_add_model_choices.py b/backend/igny8_core/modules/system/migrations/0005_add_model_choices.py
new file mode 100644
index 00000000..0b74d449
--- /dev/null
+++ b/backend/igny8_core/modules/system/migrations/0005_add_model_choices.py
@@ -0,0 +1,33 @@
+# Generated by Django 5.2.9 on 2025-12-20 14:43
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('system', '0004_fix_global_settings_remove_override'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='globalintegrationsettings',
+            name='anthropic_model',
+            field=models.CharField(choices=[('claude-3-5-sonnet-20241022', 'Claude 3.5 Sonnet (Oct 2024) - $3.00 / $15.00 per 1M tokens'), ('claude-3-5-sonnet-20240620', 'Claude 3.5 Sonnet (Jun 2024) - $3.00 / $15.00 per 1M tokens'), ('claude-3-opus-20240229', 'Claude 3 Opus - $15.00 / $75.00 per 1M tokens'), ('claude-3-sonnet-20240229', 'Claude 3 Sonnet - $3.00 / $15.00 per 1M tokens'), ('claude-3-haiku-20240307', 'Claude 3 Haiku - $0.25 / $1.25 per 1M tokens')], default='claude-3-5-sonnet-20241022', help_text='Default Anthropic model (accounts can override if plan allows)', max_length=100),
+        ),
+        migrations.AlterField(
+            model_name='globalintegrationsettings',
+            name='dalle_model',
+            field=models.CharField(choices=[('dall-e-3', 'DALL·E 3 - $0.040 per image'), ('dall-e-2', 'DALL·E 2 - $0.020 per image')], default='dall-e-3', help_text='Default DALL-E model (accounts can override if plan allows)', max_length=100),
+        ),
+        migrations.AlterField(
+            model_name='globalintegrationsettings',
+            name='dalle_size',
+            field=models.CharField(choices=[('1024x1024', '1024x1024 (Square)'), ('1792x1024', '1792x1024 (Landscape)'), ('1024x1792', '1024x1792 (Portrait)'), ('512x512', '512x512 (Small Square)')], default='1024x1024', help_text='Default image size (accounts can override if plan allows)', max_length=20),
+        ),
+        migrations.AlterField(
+            model_name='globalintegrationsettings',
+            name='openai_model',
+            field=models.CharField(choices=[('gpt-4.1', 'GPT-4.1 - $2.00 / $8.00 per 1M tokens'), ('gpt-4o-mini', 'GPT-4o mini - $0.15 / $0.60 per 1M tokens'), ('gpt-4o', 'GPT-4o - $2.50 / $10.00 per 1M tokens'), ('gpt-4-turbo-preview', 'GPT-4 Turbo Preview - $10.00 / $30.00 per 1M tokens'), ('gpt-5.1', 'GPT-5.1 - $1.25 / $10.00 per 1M tokens (16K)'), ('gpt-5.2', 'GPT-5.2 - $1.75 / $14.00 per 1M tokens (16K)')], default='gpt-4o-mini', help_text='Default text generation model (accounts can override if plan allows)', max_length=100),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/migrations/0006_fix_image_settings.py b/backend/igny8_core/modules/system/migrations/0006_fix_image_settings.py
new file mode 100644
index 00000000..e388d95e
--- /dev/null
+++ b/backend/igny8_core/modules/system/migrations/0006_fix_image_settings.py
@@ -0,0 +1,44 @@
+# Generated by Django 5.2.9 on 2025-12-20 14:51
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('system', '0005_add_model_choices'),
+    ]
+
+    operations = [
+        migrations.RemoveField(
+            model_name='globalintegrationsettings',
+            name='anthropic_api_key',
+        ),
+        migrations.RemoveField(
+            model_name='globalintegrationsettings',
+            name='anthropic_model',
+        ),
+        migrations.RemoveField(
+            model_name='globalintegrationsettings',
+            name='dalle_quality',
+        ),
+        migrations.RemoveField(
+            model_name='globalintegrationsettings',
+            name='dalle_style',
+        ),
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='image_quality',
+            field=models.CharField(choices=[('standard', 'Standard'), ('hd', 'HD')], default='standard', help_text='Default image quality for all providers (accounts can override if plan allows)', max_length=20),
+        ),
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='image_style',
+            field=models.CharField(choices=[('vivid', 'Vivid'), ('natural', 'Natural'), ('realistic', 'Realistic'), ('artistic', 'Artistic'), ('cartoon', 'Cartoon')], default='realistic', help_text='Default image style for all providers (accounts can override if plan allows)', max_length=20),
+        ),
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='runware_model',
+            field=models.CharField(choices=[('runware:97@1', 'Runware 97@1 - Versatile Model'), ('runware:100@1', 'Runware 100@1 - High Quality'), ('runware:101@1', 'Runware 101@1 - Fast Generation')], default='runware:97@1', help_text='Default Runware model (accounts can override if plan allows)', max_length=100),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/migrations/0007_add_image_defaults.py b/backend/igny8_core/modules/system/migrations/0007_add_image_defaults.py
new file mode 100644
index 00000000..070f65f0
--- /dev/null
+++ b/backend/igny8_core/modules/system/migrations/0007_add_image_defaults.py
@@ -0,0 +1,28 @@
+# Generated by Django 5.2.9 on 2025-12-20 15:05
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('system', '0006_fix_image_settings'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='desktop_image_size',
+            field=models.CharField(default='1024x1024', help_text='Default desktop image size (accounts can override if plan allows)', max_length=20),
+        ),
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='max_in_article_images',
+            field=models.IntegerField(default=2, help_text='Default maximum images to generate per article (1-5, accounts can override if plan allows)'),
+        ),
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='mobile_image_size',
+            field=models.CharField(default='512x512', help_text='Default mobile image size (accounts can override if plan allows)', max_length=20),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/migrations/0008_add_default_image_service.py b/backend/igny8_core/modules/system/migrations/0008_add_default_image_service.py
new file mode 100644
index 00000000..a99d8865
--- /dev/null
+++ b/backend/igny8_core/modules/system/migrations/0008_add_default_image_service.py
@@ -0,0 +1,18 @@
+# Generated by Django 5.2.9 on 2025-12-20 15:12
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('system', '0007_add_image_defaults'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='globalintegrationsettings',
+            name='default_image_service',
+            field=models.CharField(choices=[('openai', 'OpenAI DALL-E'), ('runware', 'Runware')], default='openai', help_text='Default image generation service for all accounts (openai=DALL-E, runware=Runware)', max_length=20),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/migrations/0009_fix_variables_optional.py b/backend/igny8_core/modules/system/migrations/0009_fix_variables_optional.py
new file mode 100644
index 00000000..3904239b
--- /dev/null
+++ b/backend/igny8_core/modules/system/migrations/0009_fix_variables_optional.py
@@ -0,0 +1,18 @@
+# Generated by Django 5.2.9 on 2025-12-20 19:33
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('system', '0008_add_default_image_service'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='globalaiprompt',
+            name='variables',
+            field=models.JSONField(blank=True, default=list, help_text='Optional: List of variables used in the prompt (e.g., {keyword}, {industry})'),
+        ),
+    ]
diff --git a/backend/igny8_core/modules/system/models.py b/backend/igny8_core/modules/system/models.py
index 734d61dc..36e2847b 100644
--- a/backend/igny8_core/modules/system/models.py
+++ b/backend/igny8_core/modules/system/models.py
@@ -84,11 +84,23 @@ class AIPrompt(AccountBaseModel):
             return None
     
     def reset_to_default(self):
-        """Reset prompt to global default"""
-        if self.default_prompt:
-            self.prompt_value = self.default_prompt
+        """Reset prompt to global default from GlobalAIPrompt"""
+        from .global_settings_models import GlobalAIPrompt
+        
+        try:
+            global_prompt = GlobalAIPrompt.objects.get(
+                prompt_type=self.prompt_type,
+                is_active=True
+            )
+            self.prompt_value = global_prompt.prompt_value
+            self.default_prompt = global_prompt.prompt_value
             self.is_customized = False
             self.save()
+        except GlobalAIPrompt.DoesNotExist:
+            raise ValueError(
+                f"Cannot reset: Global prompt '{self.prompt_type}' not found. "
+                f"Please configure it in Django admin at: /admin/system/globalaiprompt/"
+            )
     
     def __str__(self):
         status = "Custom" if self.is_customized else "Default"
diff --git a/backend/igny8_core/modules/system/utils.py b/backend/igny8_core/modules/system/utils.py
index 320f7331..35579bbd 100644
--- a/backend/igny8_core/modules/system/utils.py
+++ b/backend/igny8_core/modules/system/utils.py
@@ -5,8 +5,14 @@ from typing import Optional
 
 
 def get_default_prompt(prompt_type: str) -> str:
-    """Get default prompt value by type"""
-    defaults = {
+    """Get default prompt value from GlobalAIPrompt ONLY - single source of truth"""
+    from .global_settings_models import GlobalAIPrompt
+    
+    try:
+        global_prompt = GlobalAIPrompt.objects.get(prompt_type=prompt_type, is_active=True)
+        return global_prompt.prompt_value
+    except GlobalAIPrompt.DoesNotExist:
+        return f"ERROR: Global prompt '{prompt_type}' not configured in admin. Please configure it at: admin/system/globalaiprompt/"
         'clustering': """You are a semantic strategist and SEO architecture engine. Your task is to analyze the provided keyword list and group them into meaningful, intent-driven topic clusters that reflect how real users search, think, and act online.
 
 Return a single JSON object with a "clusters" array. Each cluster must follow this structure:
diff --git a/frontend/src/components/common/ValidationCard.tsx b/frontend/src/components/common/ValidationCard.tsx
index 2848f623..a0a61cda 100644
--- a/frontend/src/components/common/ValidationCard.tsx
+++ b/frontend/src/components/common/ValidationCard.tsx
@@ -46,33 +46,20 @@ export default function ValidationCard({
     setTestResult(null);
 
     try {
-      // Get saved settings to get API key and model
+      // Get saved settings to get model
       // fetchAPI extracts data from unified format {success: true, data: {...}}
       // So settingsData IS the data object (config object)
       const settingsData = await fetchAPI(`/v1/system/settings/integrations/${integrationId}/`);
       
-      let apiKey = '';
-      let model = 'gpt-4.1';
+      let model = 'gpt-4o-mini';
       
       if (settingsData && typeof settingsData === 'object') {
-        apiKey = settingsData.apiKey || '';
-        model = settingsData.model || 'gpt-4.1';
+        model = settingsData.model || 'gpt-4o-mini';
       }
 
-      if (!apiKey) {
-        setTestResult({
-          success: false,
-          message: 'API key not configured. Please configure your API key in settings first.',
-        });
-        setIsLoading(false);
-        return;
-      }
-
-      // Call test endpoint
+      // Call test endpoint (uses platform API key - no apiKey parameter needed)
       // For Runware, we don't need with_response or model config
-      const requestBody: any = {
-        apiKey: apiKey,
-      };
+      const requestBody: any = {};
       
       if (integrationId === 'openai') {
         requestBody.config = {
diff --git a/frontend/src/layout/AppSidebar.tsx b/frontend/src/layout/AppSidebar.tsx
index 3cf65ac4..c148dcda 100644
--- a/frontend/src/layout/AppSidebar.tsx
+++ b/frontend/src/layout/AppSidebar.tsx
@@ -17,7 +17,6 @@ import {
   FileIcon,
   UserIcon,
   UserCircleIcon,
-  BoxCubeIcon,
 } from "../icons";
 import { useSidebar } from "../context/SidebarContext";
 import SidebarWidget from "./SidebarWidget";
@@ -216,36 +215,10 @@ const AppSidebar: React.FC = () => {
             name: "Profile Settings",
             path: "/settings/profile",
           },
-          // Integration is admin-only; hide for non-privileged users (handled in render)
           {
             icon: <PlugInIcon />,
-            name: "Integration",
+            name: "AI Model Settings",
             path: "/settings/integration",
-            adminOnly: true,
-          },
-          // Global Settings - Admin only, dropdown with global config pages
-          {
-            icon: <BoxCubeIcon />,
-            name: "Global Settings",
-            adminOnly: true,
-            subItems: [
-              {
-                name: "Platform API Keys",
-                path: "/admin/system/globalintegrationsettings/",
-              },
-              {
-                name: "Global Prompts",
-                path: "/admin/system/globalaiprompt/",
-              },
-              {
-                name: "Global Author Profiles",
-                path: "/admin/system/globalauthorprofile/",
-              },
-              {
-                name: "Global Strategies",
-                path: "/admin/system/globalstrategy/",
-              },
-            ],
           },
           {
             icon: <PageIcon />,
diff --git a/frontend/src/pages/Settings/Integration.tsx b/frontend/src/pages/Settings/Integration.tsx
index daf30120..ed20db2d 100644
--- a/frontend/src/pages/Settings/Integration.tsx
+++ b/frontend/src/pages/Settings/Integration.tsx
@@ -411,23 +411,24 @@ export default function Integration() {
     if (integrationId === 'openai') {
       return [
         { label: 'App Name', value: 'OpenAI API' },
-        { label: 'API Key', value: config.apiKey ? `${config.apiKey.substring(0, 20)}...` : 'Not configured' },
-        { label: 'Model', value: config.model || 'Not set' },
+        { label: 'Model', value: config.model || 'gpt-4o-mini' },
+        { label: 'Status', value: config.using_global ? 'Using platform defaults' : 'Custom settings' },
       ];
     } else if (integrationId === 'runware') {
       return [
         { label: 'App Name', value: 'Runware API' },
-        { label: 'API Key', value: config.apiKey ? `${config.apiKey.substring(0, 20)}...` : 'Not configured' },
+        { label: 'Status', value: config.using_global ? 'Using platform defaults' : 'Custom settings' },
       ];
     } else if (integrationId === 'image_generation') {
       const service = config.service || 'openai';
       const modelDisplay = service === 'openai' 
-        ? (config.model || 'Not set')
-        : (config.runwareModel || 'Not set');
+        ? (config.model || config.imageModel || 'dall-e-3')
+        : (config.runwareModel || 'runware:97@1');
       
       return [
-        { label: 'Service', value: service === 'openai' ? 'OpenAI' : 'Runware' },
+        { label: 'Service', value: service === 'openai' ? 'OpenAI DALL-E' : 'Runware' },
         { label: 'Model', value: modelDisplay },
+        { label: 'Status', value: config.using_global ? 'Using platform defaults' : 'Custom settings' },
       ];
     }
     return [];