usage calculation

add docs, fix anthropic upstream model alias problem

README.md

@@ -99,6 +99,7 @@ The most common settings are shown below. See `.env.example` for the full list.
 - `NPUB` – Nostr public key of the proxy
 - `HTTP_URL` – Public-facing URL of the proxy
 - `ONION_URL` – Tor hidden service URL of the proxy
+- `NEXT_PUBLIC_API_URL` - UI Configuration for Next.js frontend (proxy URL, default: 'http://127.0.0.1:8000' )
 
 ## Database Migrations
 

docs/getting-started/ui-configuration.md

@@ -0,0 +1,53 @@
+# UI Configuration
+
+This guide explains how to configure the Routstr UI for different environments.
+
+## Environment Variables
+
+The UI uses Next.js environment variables to configure API endpoints and authentication.
+
+### Centralized Configuration
+
+This project uses a centralized configuration approach with a single `.env` file in the project root. This file contains both backend and frontend configuration variables.
+
+Create or update your `.env` file in the project root:
+
+```bash
+# .env (in project root)
+
+# UI Configuration (NEXT_PUBLIC_ variables are exposed to the browser)
+NEXT_PUBLIC_API_URL=http://127.0.0.1:8000
+```
+
+### Development vs Production
+
+The same `.env` file is used for both development and production. Simply change the values:
+
+**Development:**
+
+```bash
+NEXT_PUBLIC_API_URL=http://127.0.0.1:8000
+```
+
+**Production:**
+
+```bash
+NEXT_PUBLIC_API_URL=https://api.yourroutstr.com
+```
+
+## Building the UI
+
+The build process automatically reads configuration from the root `.env` file:
+
+```bash
+# From the project root
+make ui-build
+# or
+./scripts/build-ui.sh
+```
+
+The build script will automatically:
+
+- Load `NEXT_PUBLIC_*` variables from the root `.env` file
+- Use them during the Next.js build process
+- Display warnings if the `.env` file is missing

docs/user-guide/admin-dashboard.md

@@ -1,329 +1,216 @@
 # Admin Dashboard
 
-The Routstr admin dashboard provides a web interface for managing your node, viewing balances, and handling withdrawals.
+The Routstr admin dashboard is a modern web interface for managing your node, monitoring wallet balances, configuring AI models and providers, and handling Bitcoin Lightning payments through Cashu eCash.
 
 ## Accessing the Dashboard
 
-### URL Format
-
-The admin dashboard is available at:
-
-```
-https://api.routstr.com/admin/
-```
-
-> **Important**: Always include the trailing slash (`/`) in the URL.
-
 ### Authentication
 
-The dashboard is protected by a password set in the `ADMIN_PASSWORD` environment variable.
+The dashboard is protected by password authentication:
 
-1. Navigate to `/admin/`
+1. Navigate to `/admin/` in your browser
 2. Enter the admin password
-3. Click "Login"
+3. Optional: Configure custom base URL if not pre-configured
+4. Click "Login"
 
-The password is stored as a secure cookie for the session.
+The interface supports both environment-configured URLs and manual URL entry for deployment flexibility.
 
 ## Dashboard Overview
 
-### Main Interface
+The main dashboard consists of four primary sections accessible through a collapsible sidebar:
 
-The dashboard displays:
+- **Dashboard** - Wallet balance monitoring and fund management
+- **Models** - AI model management and testing
+- **Providers** - Upstream provider configuration
+- **Settings** - Node configuration and admin preferences
 
-- **Node Information**
-  - Node name and description
-  - Version number
-  - Public URLs (HTTP and Onion)
-  - Supported Cashu mints
+### Navigation
 
-- **Statistics**
-  - Total API keys
-  - Active keys
-  - Total balance across all keys
-  - Recent activity
+## Dashboard Page
 
-- **API Key List**
-  - All keys with balances
-  - Usage statistics
-  - Management options
+### Wallet Balance Management
 
-## Features
+#### Balance Display Options
 
-### Viewing API Keys
+Switch between display units using the toggle buttons:
 
-The main table shows all API keys with:
+- **msat** - Millisatoshis (highest precision)
+- **sat** - Satoshis (standard Bitcoin unit)
+- **usd** - US Dollar equivalent (when exchange rate available)
 
-| Column | Description |
-|--------|-------------|
-| API Key | Masked key (first/last 4 chars) |
-| Balance | Current balance in sats |
-| Created | Creation timestamp |
-| Last Used | Most recent API call |
-| Total Spent | Lifetime usage |
-| Status | Active/Expired/Disabled |
+#### Balance Overview
 
-### Searching and Filtering
+The dashboard displays three key metrics:
 
-- **Search**: Find keys by partial match
-- **Sort**: Click column headers to sort
-- **Filter**: Show only active/expired keys
-- **Export**: Download data as CSV
+- **Your Balance (Total)** - Available funds for node operator
+- **Total Wallet** - Combined balance across all Cashu mints
+- **User Balance** - Funds held for API key holders
 
-### Key Details
+#### Detailed Balance Breakdown
 
-Click on any key to view:
+View balances by mint with the following information:
 
-- Full API key (masked by default)
-- Complete transaction history
-- Usage graphs
-- Metadata (name, expiry, refund address)
+| Column      | Description                           |
+| ----------- | ------------------------------------- |
+| Mint / Unit | Cashu mint URL and currency unit      |
+| Wallet      | Total funds in this mint              |
+| Users       | Funds belonging to API key holders    |
+| Owner       | Your available funds (Wallet - Users) |
 
-## Balance Management
+### Temporary Balances
 
-### Viewing Balances
+Monitor API key activity with:
 
-Balances are displayed in multiple units:
+- **Summary Cards** - Total balance, total spent, total requests
+- **Search Functionality** - Filter by key hash or refund address
+- **Detailed Table** - Individual key balances with expiry times
+- **Auto-refresh** - Updates every 60 seconds
 
-- **Sats**: Standard satoshi units
-- **mSats**: Millisatoshis (internal precision)
-- **BTC**: Bitcoin decimal format
-- **USD**: Approximate USD value
+### Fund Management
 
-### Balance History
+#### Withdrawing Funds
 
-View balance changes over time:
+To withdraw your available balance:
 
-```
-Time         | Type      | Amount  | Balance | Description
--------------|-----------|---------|---------|-------------
-12:34:56     | Deposit   | +10,000 | 10,000  | Token redemption
-12:35:12     | Usage     | -154    | 9,846   | gpt-3.5-turbo call
-12:36:45     | Usage     | -210    | 9,636   | gpt-4 call
-```
+1. Click the **Withdraw** button
+2. Select which mint to withdraw from
+3. Specify the amount (or withdraw full balance)
+4. Click **Generate Token**
+5. Copy the generated eCash token
+6. Import the token into your Cashu wallet
 
-## Withdrawals
+#### Real-time Updates
 
-### Manual Withdrawal
+- Balances refresh automatically every 30 seconds
+- Manual refresh option available
+- Live Bitcoin/USD exchange rate integration
+- Error handling for mint connectivity issues
 
-To withdraw funds from an API key:
+## Models Management Page
 
-1. Click "Withdraw" next to the key
-2. Optionally specify amount (default: full balance)
-3. Select target Cashu mint
-4. Click "Generate Token"
-5. Copy the eCash token
-6. Redeem in your Cashu wallet
+### Model Organization
 
-### Bulk Operations
+Models are organized by provider groups with tabs:
 
-For multiple withdrawals:
+- **All Models** - Combined view of all available models
+- **Provider-specific tabs** - Individual providers (OpenRouter, Azure, etc.)
+- Badge indicators showing active/total model counts
 
-1. Select keys using checkboxes
-2. Click "Bulk Actions" → "Withdraw"
-3. Tokens are generated for each key
-4. Download all tokens as text file
+### Model Management Features
 
-### Automatic Withdrawals
+#### Individual Model Operations
 
-If configured with `RECEIVE_LN_ADDRESS`:
+For each model you can:
 
-- Balances above threshold auto-convert to Lightning
-- Sent to configured Lightning address
-- View payout history in dashboard
+- **Toggle Enable/Disable** - Control model availability
+- **View Details** - Context length, pricing, description
+- **Edit Configuration** - Model-specific settings
+- **Status Indicators** - Green badges for enabled, gray for disabled
 
-## Node Configuration
+#### Bulk Operations
 
-### Viewing Settings
+- **Select All/Deselect All** - Quick selection controls
+- **Bulk Enable/Disable** - Mass model management
+- **Bulk Delete** - Remove model overrides
+- **Provider-level Actions** - Apply settings to all models in a provider
 
-Current node configuration is displayed:
+#### Model Information Display
 
-- Upstream provider URL
-- Enabled features
-- Pricing model
-- Fee structure
+- **Model Types** - Text, embedding, image, audio, multimodal indicators
+- **Pricing Information** - Per-million-token costs for input/output
+- **Context Length** - Maximum tokens supported
+- **API Key Status** - Whether credentials are configured
+- **Free Model Indicators** - No-cost models clearly marked
 
-### Models and Pricing
+## Providers Management Page
 
-View supported models and their pricing:
+### Upstream Provider Configuration
 
-| Model | Input $/1K | Output $/1K | Sats/1K |
-|-------|------------|-------------|---------|
-| gpt-3.5-turbo | $0.0015 | $0.002 | 3/4 |
-| gpt-4 | $0.03 | $0.06 | 60/120 |
-| dall-e-3 | - | - | 1000/image |
+Manage AI provider connections and credentials:
 
-### Updating Configuration
+#### Provider Types Supported
 
-> **Note**: Configuration changes require node restart.
+- **OpenRouter** - Multi-model aggregator
+- **Azure OpenAI** - Microsoft's OpenAI service
+- **OpenAI** - Direct OpenAI integration
+- **Custom Providers** - Any OpenAI-compatible API
 
-To update settings:
+#### Adding New Providers
 
-1. Modify environment variables
-2. Restart the node
-3. Verify changes in dashboard
+1. Click **Add Provider**
+2. Select **Provider Type** from dropdown
+3. Enter **Base URL** (auto-populated for known providers)
+4. Add **API Key** for authentication
+5. Set **API Version** (required for Azure)
+6. Toggle **Enabled** status
+7. Click **Create**
 
-## Analytics
+#### Provider Management
 
-### Usage Statistics
+**Provider Cards Display:**
 
-View comprehensive usage data:
+- Provider type and status (Enabled/Disabled)
+- Base URL configuration
+- Action buttons (Models, Edit, Delete)
 
-- **Requests per Day**: Line graph
-- **Token Usage**: Stacked bar chart
-- **Model Distribution**: Pie chart
-- **Cost Analysis**: Breakdown by model
+**Available Actions:**
 
-### Performance Metrics
+- **Edit** - Modify provider configuration
+- **Delete** - Remove provider (with confirmation)
+- **View Models** - Expand model discovery interface
+- **Enable/Disable** - Toggle provider availability
 
-Monitor node performance:
+#### Model Discovery
 
-- Average response time
-- Request success rate
-- Upstream API latency
-- Cache hit ratio
+Each provider shows two types of models:
 
-### Export Data
+**Provided Models Tab:**
 
-Export analytics data:
+- Auto-discovered from provider's catalog
+- Read-only model information
+- Real-time availability updates
 
-1. Select date range
-2. Choose metrics
-3. Click "Export"
-4. Download as CSV/JSON
+**Custom Models Tab:**
 
-## Security Features
+- Manually configured model overrides
+- Extend or override provider catalog
+- Individual enable/disable controls
 
-### Access Control
+## Settings Page
 
-- Password protection
-- Session timeout (configurable)
-- IP allowlisting (optional)
-- Audit logging
+### Node Configuration
 
-### Security Log
+Configure core node settings and preferences:
 
-View security events:
+#### Basic Information
 
-```
-2024-01-15 12:34:56 | Login Success | IP: 192.168.1.1
-2024-01-15 12:35:12 | Withdrawal | Key: sk-****abcd | Amount: 5000
-2024-01-15 12:40:00 | Session Timeout | IP: 192.168.1.1
-```
+- **Node Name** - Identifier for your node
+- **Node Description** - Descriptive text for your service
+- **HTTP URL** - Public HTTP endpoint
+- **Onion URL** - Tor hidden service address
 
-### Best Practices
+#### Nostr Integration
 
-1. **Strong Password**: Use a long, random password
-2. **HTTPS Only**: Always access via HTTPS
-3. **Regular Monitoring**: Check logs frequently
-4. **Limited Access**: Restrict dashboard access
+- **Public Key (npub)** - Your Nostr public identity
+- **Private Key (nsec)** - Nostr private key with show/hide toggle
+- **Nostr Relays** - Configure relays for provider announcements
 
-## Troubleshooting
+#### Cashu Mint Management
 
-### Cannot Access Dashboard
+- **Add Mint URLs** - Configure multiple Cashu mint endpoints
+- **Remove Mints** - Delete unused mint configurations
+- **Mint Validation** - Verify mint endpoint connectivity
 
-**Issue**: 404 Not Found
+#### Settings Features
 
-- Ensure trailing slash: `/admin/`
-- Check if admin routes are enabled
-
-**Issue**: Unauthorized
-
-- Verify `ADMIN_PASSWORD` is set
-- Clear browser cookies
-- Try incognito/private mode
-
-### Display Issues
-
-**Issue**: Broken Layout
-
-- Clear browser cache
-- Disable ad blockers
-- Try different browser
-
-**Issue**: Missing Data
-
-- Check database connectivity
-- Verify node is running
-- Review error logs
-
-### Withdrawal Problems
-
-**Issue**: Token Generation Fails
-
-- Check mint connectivity
-- Verify sufficient balance
-- Try different mint
-
-**Issue**: Invalid Token
-
-- Ensure complete token copy
-- Check token hasn't expired
-- Verify mint compatibility
-
-## Advanced Features
-
-### Custom Branding
-
-Customize dashboard appearance:
-
-```bash
-# Environment variables
-ADMIN_LOGO_URL=https://example.com/logo.png
-ADMIN_THEME_COLOR=#FF6B00
-ADMIN_CUSTOM_CSS=/path/to/custom.css
-```
-
-### API Access
-
-Access admin functions programmatically:
-
-```bash
-# Get node stats
-curl -X GET https://your-node.com/admin/api/stats \
-  -H "X-Admin-Password: your-password"
-
-# Export key data
-curl -X GET https://your-node.com/admin/api/keys \
-  -H "X-Admin-Password: your-password" \
-  -H "Accept: application/json"
-```
-
-### Webhooks
-
-Configure notifications:
-
-```bash
-ADMIN_WEBHOOK_URL=https://example.com/webhook
-ADMIN_WEBHOOK_EVENTS=withdrawal,low_balance,error
-```
-
-## Dashboard Shortcuts
-
-### Keyboard Navigation
-
-- `Ctrl+K`: Quick search
-- `Ctrl+R`: Refresh data
-- `Ctrl+E`: Export current view
-- `Escape`: Close modals
-
-### Quick Actions
-
-- Double-click to copy API key
-- Right-click for context menu
-- Drag to reorder columns
-- Shift-click to select multiple
-
-## Mobile Access
-
-The dashboard is mobile-responsive:
-
-- Touch-optimized controls
-- Swipe navigation
-- Compact view mode
-- Offline capability
+- **Real-time Save** - Changes apply immediately
+- **Validation** - Form validation with error feedback
+- **Secure Fields** - Password masking with reveal toggles
+- **Reload Functionality** - Refresh configuration from server
 
 ## Next Steps
 
-- [Models & Pricing](models-pricing.md) - Configure pricing
-- [API Reference](../api/overview.md) - Admin API endpoints
-- [Advanced Configuration](../advanced/custom-pricing.md) - Advanced settings
+- [Payment Flow](payment-flow.md) - Understanding Bitcoin payment processing
+- [Using the API](using-api.md) - Making API requests to your node
+- [Models & Pricing](models-pricing.md) - Configuring model pricing and fees
+- [API Reference](../api/overview.md) - Complete API documentation

routstr/algorithm.py

@@ -6,7 +6,7 @@ from .core.logging import get_logger
 
 if TYPE_CHECKING:
     from .payment.models import Model
-    from .upstream import UpstreamProvider
+    from .upstream import BaseUpstreamProvider
 
 logger = get_logger(__name__)
 
@@ -59,7 +59,7 @@ def calculate_model_cost_score(model: "Model") -> float:
     return total_cost
 
 
-def get_provider_penalty(provider: "UpstreamProvider") -> float:
+def get_provider_penalty(provider: "BaseUpstreamProvider") -> float:
     """Calculate a penalty multiplier for certain providers.
 
     This allows applying policy-based adjustments beyond pure cost.
@@ -86,9 +86,9 @@ def get_provider_penalty(provider: "UpstreamProvider") -> float:
 
 def should_prefer_model(
     candidate_model: "Model",
-    candidate_provider: "UpstreamProvider",
+    candidate_provider: "BaseUpstreamProvider",
     current_model: "Model",
-    current_provider: "UpstreamProvider",
+    current_provider: "BaseUpstreamProvider",
     alias: str,
 ) -> bool:
     """Determine if candidate model should replace current model for an alias.
@@ -166,10 +166,10 @@ def should_prefer_model(
 
 
 def create_model_mappings(
-    upstreams: list["UpstreamProvider"],
+    upstreams: list["BaseUpstreamProvider"],
     overrides_by_id: dict[str, tuple],
     disabled_model_ids: set[str],
-) -> tuple[dict[str, "Model"], dict[str, "UpstreamProvider"], dict[str, "Model"]]:
+) -> tuple[dict[str, "Model"], dict[str, "BaseUpstreamProvider"], dict[str, "Model"]]:
     """Create optimal model mappings based on cost and provider preferences.
 
     This is the main entry point for the algorithm. It processes all upstream providers
@@ -196,12 +196,12 @@ def create_model_mappings(
     from .upstream import resolve_model_alias
 
     model_instances: dict[str, "Model"] = {}
-    provider_map: dict[str, "UpstreamProvider"] = {}
+    provider_map: dict[str, "BaseUpstreamProvider"] = {}
     unique_models: dict[str, "Model"] = {}
 
     # Separate OpenRouter from other providers
-    openrouter: "UpstreamProvider" | None = None
-    other_upstreams: list["UpstreamProvider"] = []
+    openrouter: "BaseUpstreamProvider" | None = None
+    other_upstreams: list["BaseUpstreamProvider"] = []
 
     for upstream in upstreams:
         base_url = getattr(upstream, "base_url", "")
@@ -215,7 +215,7 @@ def create_model_mappings(
         return model_id.split("/", 1)[1] if "/" in model_id else model_id
 
     def _maybe_set_alias(
-        alias: str, model: "Model", provider: "UpstreamProvider"
+        alias: str, model: "Model", provider: "BaseUpstreamProvider"
     ) -> None:
         """Set alias to model/provider if not set or if new model is preferred."""
         existing_model = model_instances.get(alias)
@@ -233,7 +233,7 @@ def create_model_mappings(
                 provider_map[alias] = provider
 
     def process_provider_models(
-        upstream: "UpstreamProvider", is_openrouter: bool = False
+        upstream: "BaseUpstreamProvider", is_openrouter: bool = False
     ) -> None:
         """Process all models from a given provider."""
         upstream_prefix = getattr(upstream, "upstream_name", None)
@@ -258,7 +258,11 @@ def create_model_mappings(
                 unique_models[base_id] = unique_model
 
             # Get all aliases for this model
-            aliases = resolve_model_alias(model_to_use.id, model_to_use.canonical_slug)
+            aliases = resolve_model_alias(
+                model_to_use.id,
+                model_to_use.canonical_slug,
+                alias_ids=model_to_use.alias_ids,
+            )
 
             # Add prefixed alias if applicable
             if upstream_prefix and "/" not in model_to_use.id:

routstr/payment/cost_caculation.py

@@ -1,4 +1,7 @@
+import base64
 import math
+import struct
+from typing import Optional
 
 from pydantic.v1 import BaseModel
 
@@ -25,8 +28,163 @@ class CostDataError(BaseModel):
     code: str
 
 
-async def calculate_cost(
-    response_data: dict, max_cost: int, session: AsyncSession
+def get_image_resolution_from_data_url(data_url: str) -> Optional[tuple[int, int]]:
+    """
+    Extract image resolution (width, height) from a base64 data URL without DOM.
+    Supports PNG and JPEG. Returns None if format unsupported or parsing fails.
+    """
+    try:
+        if not isinstance(data_url, str) or not data_url.startswith("data:"):
+            return None
+
+        comma_idx = data_url.find(",")
+        if comma_idx == -1:
+            return None
+
+        meta = data_url[5:comma_idx]  # e.g. "image/png;base64"
+        base64_data = data_url[comma_idx + 1:]
+
+        # Decode base64 to binary
+        try:
+            binary_data = base64.b64decode(base64_data)
+        except Exception:
+            return None
+
+        is_png = "image/png" in meta
+        is_jpeg = "image/jpeg" in meta or "image/jpg" in meta
+
+        # PNG: width/height are 4-byte big-endian at offsets 16 and 20
+        if is_png:
+            # Validate PNG signature
+            png_sig = b'\x89PNG\r\n\x1a\n'
+            if not binary_data.startswith(png_sig):
+                return None
+
+            if len(binary_data) < 24:
+                return None
+
+            # Width and height are at bytes 16-19 and 20-23 respectively
+            width = struct.unpack('>I', binary_data[16:20])[0]
+            height = struct.unpack('>I', binary_data[20:24])[0]
+
+            if width > 0 and height > 0:
+                return (width, height)
+            return None
+
+        # JPEG: parse markers to SOF0/SOF2 for dimensions
+        if is_jpeg:
+            offset = 0
+            # JPEG SOI 0xFFD8
+            if len(binary_data) < 2 or binary_data[0] != 0xFF or binary_data[1] != 0xD8:
+                return None
+            offset = 2
+
+            while offset < len(binary_data):
+                # Find marker
+                while offset < len(binary_data) and binary_data[offset] != 0xFF:
+                    offset += 1
+                if offset + 1 >= len(binary_data):
+                    break
+
+                # Skip fill bytes 0xFF
+                while offset < len(binary_data) and binary_data[offset] == 0xFF:
+                    offset += 1
+                if offset >= len(binary_data):
+                    break
+
+                marker = binary_data[offset]
+                offset += 1
+
+                # Standalone markers without length
+                if marker == 0xD8 or marker == 0xD9:  # SOI/EOI
+                    continue
+
+                if offset + 1 >= len(binary_data):
+                    break
+
+                length = (binary_data[offset] << 8) | binary_data[offset + 1]
+                offset += 2
+
+                # SOF0 (0xC0) or SOF2 (0xC2) contain dimensions
+                if marker == 0xC0 or marker == 0xC2:
+                    if length < 7 or offset + length - 2 > len(binary_data):
+                        return None
+                    # Skip precision byte
+                    height = (binary_data[offset + 1] << 8) | binary_data[offset + 2]
+                    width = (binary_data[offset + 3] << 8) | binary_data[offset + 4]
+                    if width > 0 and height > 0:
+                        return (width, height)
+                    return None
+                else:
+                    # Skip this segment
+                    offset += length - 2
+
+            return None
+
+        # Unsupported formats (e.g., webp/gif) - skip for now
+        return None
+    except Exception:
+        return None
+
+
+def calculate_image_tokens_from_messages(messages: list) -> int:
+    """
+    Calculate image tokens from messages using 32px patch method.
+    """
+    image_tokens = 0
+
+    try:
+        for msg in messages:
+            content = msg.get("content")
+            if isinstance(content, list):
+                for part in content:
+                    if (isinstance(part, dict) and
+                        part.get("type") == "image_url"):
+
+                        image_url = part.get("image_url")
+                        url: Optional[str] = None
+                        if isinstance(image_url, str):
+                            url = image_url
+                        elif isinstance(image_url, dict):
+                            url = image_url.get("url")
+                        else:
+                            continue
+
+                        # Expecting a base64 data URL for local image inputs
+                        if url and isinstance(url, str) and url.startswith("data:"):
+                            resolution = get_image_resolution_from_data_url(url)
+                            if resolution:
+                                width, height = resolution
+                                patch_size = 32
+                                patches_w = (width + patch_size - 1) // patch_size
+                                patches_h = (height + patch_size - 1) // patch_size
+                                tokens_from_image = patches_w * patches_h
+                                image_tokens += tokens_from_image
+
+                                logger.debug(
+                                    "Calculated image tokens",
+                                    extra={
+                                        "width": width,
+                                        "height": height,
+                                        "tokens_from_image": tokens_from_image,
+                                    }
+                                )
+                            else:
+                                logger.warning(
+                                    "Could not determine image resolution",
+                                    extra={"url_prefix": url[:50] + "..." if len(url) > 50 else url}
+                                )
+    except Exception as e:
+        logger.error(
+            "Error calculating image tokens",
+            extra={"error": str(e), "error_type": type(e).__name__}
+        )
+
+    return image_tokens
+
+
+async def calculate_cost(  # todo: can be sync
+    response_data: dict, max_cost: int, session: AsyncSession, request_data: Optional[dict] = None
 ) -> CostData | MaxCostData | CostDataError:
     """
     Calculate the cost of an API request based on token usage.
@@ -34,6 +192,8 @@ async def calculate_cost(
     Args:
         response_data: Response data containing usage information
         max_cost: Maximum cost in millisats
+        session: Database session
+        request_data: Original request data containing messages (for image token calculation)
 
     Returns:
         Cost data or error information
@@ -78,13 +238,9 @@ async def calculate_cost(
             extra={"model": response_model},
         )
 
-        from ..proxy import get_upstreams
-        from ..upstream import get_model_with_override
+        from ..proxy import get_model_instance
 
-        upstreams = get_upstreams()
-        model_obj = await get_model_with_override(
-            response_model, upstreams, session=session
-        )
+        model_obj = get_model_instance(response_model)
 
         if not model_obj:
             logger.error(
@@ -136,14 +292,28 @@ async def calculate_cost(
     input_tokens = response_data.get("usage", {}).get("prompt_tokens", 0)
     output_tokens = response_data.get("usage", {}).get("completion_tokens", 0)
 
-    input_msats = round(input_tokens / 1000 * MSATS_PER_1K_INPUT_TOKENS, 3)
+    # Calculate image tokens from request data if available
+    image_tokens = 0
+    if request_data and "messages" in request_data:
+        image_tokens = calculate_image_tokens_from_messages(request_data["messages"])
+        logger.debug(
+            "Image tokens calculated",
+            extra={"image_tokens": image_tokens, "text_input_tokens": input_tokens}
+        )
+
+    # Add image tokens to input tokens for cost calculation
+    total_input_tokens = input_tokens + image_tokens
+
+    input_msats = round(total_input_tokens / 1000 * MSATS_PER_1K_INPUT_TOKENS, 3)
     output_msats = round(output_tokens / 1000 * MSATS_PER_1K_OUTPUT_TOKENS, 3)
     token_based_cost = math.ceil(input_msats + output_msats)
 
     logger.info(
         "Calculated token-based cost",
         extra={
-            "input_tokens": input_tokens,
+            "text_input_tokens": input_tokens,
+            "image_tokens": image_tokens,
+            "total_input_tokens": total_input_tokens,
             "output_tokens": output_tokens,
             "input_cost_msats": input_msats,
             "output_cost_msats": output_msats,

routstr/payment/helpers.py

@@ -104,11 +104,9 @@ async def get_max_cost_for_model(
         return max(settings.min_request_msat, default_cost_msats)
 
     if not model_obj:
-        from ..proxy import get_upstreams
-        from ..upstream import get_model_with_override
+        from ..proxy import get_model_instance
 
-        upstreams = get_upstreams()
-        model_obj = await get_model_with_override(model, upstreams, session)
+        model_obj = get_model_instance(model)
 
     if not model_obj:
         fallback_msats = settings.fixed_cost_per_request * 1000

routstr/payment/models.py

@@ -60,6 +60,7 @@ class Model(BaseModel):
     enabled: bool = True
     upstream_provider_id: int | None = None
     canonical_slug: str | None = None
+    alias_ids: list[str] | None = None
 
     def __hash__(self) -> int:
         return hash(self.id)
@@ -409,6 +410,7 @@ def _update_model_sats_pricing(model: Model, sats_to_usd: float) -> Model:
             enabled=model.enabled,
             upstream_provider_id=model.upstream_provider_id,
             canonical_slug=model.canonical_slug,
+            alias_ids=model.alias_ids,
         )
     except Exception as e:
         logger.error(
@@ -577,7 +579,6 @@ async def _cleanup_enabled_models_once() -> None:
 
         for db_model in db_models:
             # Find corresponding upstream model
-            print(db_model.id)
             upstream_model = None
             for upstream in upstreams:
                 upstream_model = upstream.get_cached_model_by_id(db_model.id)
@@ -613,7 +614,7 @@ async def _cleanup_enabled_models_once() -> None:
 
 
 def _pricing_matches(
-    db_pricing: dict, upstream_pricing: dict, tolerance: float = 0.1
+    db_pricing: dict, upstream_pricing: dict, tolerance: float = 0.0
 ) -> bool:
     """Check if pricing dictionaries match within tolerance."""
     keys_to_compare = [
@@ -626,9 +627,8 @@ def _pricing_matches(
     ]
 
     for key in keys_to_compare:
-        db_val = float(db_pricing.get(key, 0.0)) * 1000000
-        upstream_val = float(upstream_pricing.get(key, 0.0)) * 1000000
-        print(db_val - upstream_val)
+        db_val = int(float(db_pricing.get(key, 0.0)) * 1000000)
+        upstream_val = int(float(upstream_pricing.get(key, 0.0)) * 1000000)
 
         if abs(db_val - upstream_val) > tolerance:
             return False

routstr/proxy.py

@@ -3,7 +3,7 @@ from typing import Any
 
 from fastapi import APIRouter, Depends, HTTPException, Request
 from fastapi.responses import Response, StreamingResponse
-from sqlmodel import col, select
+from sqlmodel import select
 
 from .algorithm import create_model_mappings
 from .auth import pay_for_request, revert_pay_for_request, validate_bearer_key
@@ -23,14 +23,14 @@ from .payment.helpers import (
     get_max_cost_for_model,
 )
 from .payment.models import Model
-from .upstream import UpstreamProvider, init_upstreams
+from .upstream import BaseUpstreamProvider, init_upstreams
 
 logger = get_logger(__name__)
 proxy_router = APIRouter()
 
-_upstreams: list[UpstreamProvider] = []
+_upstreams: list[BaseUpstreamProvider] = []
 _model_instances: dict[str, Model] = {}  # All aliases -> Model
-_provider_map: dict[str, UpstreamProvider] = {}  # All aliases -> Provider
+_provider_map: dict[str, BaseUpstreamProvider] = {}  # All aliases -> Provider
 _unique_models: dict[str, Model] = {}  # Unique model.id -> Model (no duplicates)
 
 
@@ -53,7 +53,7 @@ async def reinitialize_upstreams() -> None:
     await refresh_model_maps()
 
 
-def get_upstreams() -> list[UpstreamProvider]:
+def get_upstreams() -> list[BaseUpstreamProvider]:
     """Get the initialized upstream providers.
 
     Returns:
@@ -67,7 +67,7 @@ def get_model_instance(model_id: str) -> Model | None:
     return _model_instances.get(model_id)
 
 
-def get_provider_for_model(model_id: str) -> UpstreamProvider | None:
+def get_provider_for_model(model_id: str) -> BaseUpstreamProvider | None:
     """Get UpstreamProvider for model ID from global cache."""
     return _provider_map.get(model_id)
 
@@ -83,9 +83,7 @@ async def refresh_model_maps() -> None:
 
     # Gather database overrides and disabled models
     async with create_session() as session:
-        result = await session.exec(
-            select(ModelRow).where(col(ModelRow.enabled).is_(True))
-        )
+        result = await session.exec(select(ModelRow).where(ModelRow.enabled))
         override_rows = result.all()
 
         provider_result = await session.exec(select(UpstreamProviderRow))
@@ -103,7 +101,7 @@ async def refresh_model_maps() -> None:
         }
 
         disabled_result = await session.exec(
-            select(ModelRow.id).where(col(ModelRow.enabled).is_(False))
+            select(ModelRow.id).where(ModelRow.enabled == False)  # noqa: E712
         )
         disabled_model_ids = {row for row in disabled_result.all()}
 

routstr/upstream/__init__.py

@@ -0,0 +1,34 @@
+from .anthropic import AnthropicUpstreamProvider
+from .azure import AzureUpstreamProvider
+from .base import BaseUpstreamProvider
+from .generic import GenericUpstreamProvider
+from .helpers import (
+    _instantiate_provider,
+    _seed_providers_from_settings,
+    get_all_models_with_overrides,
+    init_upstreams,
+    refresh_upstreams_models_periodically,
+    resolve_model_alias,
+)
+from .ollama import OllamaUpstreamProvider
+from .openai import OpenAIUpstreamProvider
+from .openrouter import OpenRouterUpstreamProvider
+
+__all__ = [
+    # upstreams
+    "AnthropicUpstreamProvider",
+    "AzureUpstreamProvider",
+    "BaseUpstreamProvider",
+    "GenericUpstreamProvider",
+    "OllamaUpstreamProvider",
+    "OpenAIUpstreamProvider",
+    "OpenRouterUpstreamProvider",
+    # helpers
+    "resolve_model_alias",
+    "get_all_models_with_overrides",
+    "get_model_with_override",
+    "refresh_upstreams_models_periodically",
+    "init_upstreams",
+    "_seed_providers_from_settings",
+    "_instantiate_provider",
+]

routstr/upstream/anthropic.py

@@ -0,0 +1,43 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class AnthropicUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for Anthropic API."""
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        self.upstream_name = "anthropic"
+        super().__init__(
+            base_url="https://api.anthropic.com/v1",
+            api_key=api_key,
+            provider_fee=provider_fee,
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'anthropic/' prefix for Anthropic API compatibility and transform model names."""
+        if model_id.startswith("anthropic/"):
+            model_id = model_id[len("anthropic/") :]
+        fixed_transforms = {
+            "claude-haiku-4.5": "claude-haiku-4-5-20251001",
+            "claude-sonnet-4.5": "claude-sonnet-4-5-20250929",
+            "claude-opus-4.1": "claude-opus-4-1-20250805",
+            "claude-opus-4": "claude-opus-4-20250514",
+            "claude-sonnet-4": "claude-sonnet-4-20250514",
+            "claude-3.5-haiku": "claude-3-5-haiku-20241022",
+            "claude-3-haiku": "claude-3-haiku-20240307",
+            "claude-haiku-4-5": "claude-haiku-4-5-20251001",
+            "claude-sonnet-4-5": "claude-sonnet-4-5-20250929",
+            "claude-opus-4-1": "claude-opus-4-1-20250805",
+            "claude-3-5-haiku": "claude-3-5-haiku-20241022",
+        }
+        if model_id in fixed_transforms:
+            model_id = fixed_transforms[model_id]
+        return model_id
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch Anthropic models from OpenRouter API filtered by anthropic source."""
+        models_data = await async_fetch_openrouter_models(source_filter="anthropic")
+        models = [Model(**model) for model in models_data]  # type: ignore
+        for model in models:
+            model.alias_ids = [self.transform_model_name(model.id)]
+        return models

routstr/upstream/azure.py

@@ -0,0 +1,46 @@
+from typing import Mapping
+
+from .base import BaseUpstreamProvider
+
+
+class AzureUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for Azure OpenAI Service."""
+
+    def __init__(
+        self,
+        base_url: str,
+        api_key: str,
+        api_version: str,
+        provider_fee: float = 1.01,
+    ):
+        """Initialize Azure provider with API key and version.
+
+        Args:
+            base_url: Azure OpenAI endpoint base URL
+            api_key: Azure OpenAI API key for authentication
+            api_version: Azure OpenAI API version (e.g., "2024-02-15-preview")
+            provider_fee: Provider fee multiplier (default 1.01 for 1% fee)
+        """
+        super().__init__(
+            base_url=base_url,
+            api_key=api_key,
+            provider_fee=provider_fee,
+        )
+        self.api_version = api_version
+
+    def prepare_params(
+        self, path: str, query_params: Mapping[str, str] | None
+    ) -> Mapping[str, str]:
+        """Prepare query parameters for Azure OpenAI, adding API version.
+
+        Args:
+            path: Request path
+            query_params: Original query parameters from the client
+
+        Returns:
+            Query parameters dict with Azure API version added for chat completions
+        """
+        params = dict(query_params or {})
+        if path.endswith("chat/completions"):
+            params["api-version"] = self.api_version
+        return params

routstr/upstream/base.py

@@ -4,7 +4,7 @@ import json
 import re
 import traceback
 from collections.abc import AsyncGenerator
-from typing import Mapping
+from typing import Mapping, Optional
 
 import httpx
 from fastapi import BackgroundTasks, HTTPException, Request
@@ -25,7 +25,6 @@ from ..payment.models import (
     Pricing,
     _calculate_usd_max_costs,
     _update_model_sats_pricing,
-    async_fetch_openrouter_models,
 )
 from ..payment.price import sats_usd_price
 from ..wallet import recieve_token, send_token
@@ -33,7 +32,7 @@ from ..wallet import recieve_token, send_token
 logger = get_logger(__name__)
 
 
-class UpstreamProvider:
+class BaseUpstreamProvider:
     """Provider for forwarding requests to an upstream AI service API."""
 
     base_url: str
@@ -876,13 +875,14 @@ class UpstreamProvider:
                 )
 
     async def get_x_cashu_cost(
-        self, response_data: dict, max_cost_for_model: int
+        self, response_data: dict, max_cost_for_model: int, request_data: Optional[dict] = None
     ) -> MaxCostData | CostData | None:
         """Calculate cost for X-Cashu payment based on response data.
 
         Args:
             response_data: Response data containing model and usage information
             max_cost_for_model: Maximum cost for the model
+            request_data: Original request data containing messages (for image token calculation)
 
         Returns:
             Cost data object (MaxCostData or CostData) or None if calculation fails
@@ -894,7 +894,7 @@ class UpstreamProvider:
         )
 
         async with create_session() as session:
-            match await calculate_cost(response_data, max_cost_for_model, session):
+            match await calculate_cost(response_data, max_cost_for_model, session, request_data):
                 case MaxCostData() as cost:
                     logger.debug(
                         "Using max cost pricing",
@@ -1018,6 +1018,7 @@ class UpstreamProvider:
         unit: str,
         max_cost_for_model: int,
         mint: str | None = None,
+        request_data: Optional[dict] = None,
     ) -> StreamingResponse:
         """Handle streaming response for X-Cashu payment, calculating refund if needed.
 
@@ -1076,7 +1077,7 @@ class UpstreamProvider:
             response_data = {"usage": usage_data, "model": model}
             try:
                 cost_data = await self.get_x_cashu_cost(
-                    response_data, max_cost_for_model
+                    response_data, max_cost_for_model, request_data
                 )
                 if cost_data:
                     if unit == "msat":
@@ -1151,6 +1152,7 @@ class UpstreamProvider:
         unit: str,
         max_cost_for_model: int,
         mint: str | None = None,
+        request_data: Optional[dict] = None,
     ) -> Response:
         """Handle non-streaming response for X-Cashu payment, calculating refund if needed.
 
@@ -1171,7 +1173,7 @@ class UpstreamProvider:
 
         try:
             response_json = json.loads(content_str)
-            cost_data = await self.get_x_cashu_cost(response_json, max_cost_for_model)
+            cost_data = await self.get_x_cashu_cost(response_json, max_cost_for_model, request_data)
 
             if not cost_data:
                 logger.error(
@@ -1281,6 +1283,7 @@ class UpstreamProvider:
         unit: str,
         max_cost_for_model: int,
         mint: str | None = None,
+        request_data: Optional[dict] = None,
     ) -> StreamingResponse | Response:
         """Handle chat completion response for X-Cashu payment, detecting streaming vs non-streaming.
 
@@ -1317,11 +1320,11 @@ class UpstreamProvider:
 
             if is_streaming:
                 return await self.handle_x_cashu_streaming_response(
-                    content_str, response, amount, unit, max_cost_for_model, mint
+                    content_str, response, amount, unit, max_cost_for_model, mint, request_data
                 )
             else:
                 return await self.handle_x_cashu_non_streaming_response(
-                    content_str, response, amount, unit, max_cost_for_model, mint
+                    content_str, response, amount, unit, max_cost_for_model, mint, request_data
                 )
 
         except Exception as e:
@@ -1373,6 +1376,17 @@ class UpstreamProvider:
         request_body = await request.body()
         transformed_body = self.prepare_request_body(request_body, model_obj)
 
+        # Parse request data for image token calculation
+        request_data = None
+        try:
+            if request_body:
+                request_data = json.loads(request_body.decode('utf-8'))
+        except Exception as e:
+            logger.warning(
+                "Could not parse request body for image token calculation",
+                extra={"error": str(e)}
+            )
+
         logger.debug(
             "Forwarding request to upstream",
             extra={
@@ -1458,7 +1472,7 @@ class UpstreamProvider:
                     )
 
                     result = await self.handle_x_cashu_chat_completion(
-                        response, amount, unit, max_cost_for_model, mint
+                        response, amount, unit, max_cost_for_model, mint, request_data
                     )
                     background_tasks = BackgroundTasks()
                     background_tasks.add_task(response.aclose)
@@ -1626,6 +1640,7 @@ class UpstreamProvider:
             enabled=model.enabled,
             upstream_provider_id=model.upstream_provider_id,
             canonical_slug=model.canonical_slug,
+            alias_ids=model.alias_ids,
         )
 
         (
@@ -1648,6 +1663,7 @@ class UpstreamProvider:
             enabled=model.enabled,
             upstream_provider_id=model.upstream_provider_id,
             canonical_slug=model.canonical_slug,
+            alias_ids=model.alias_ids,
         )
 
     async def fetch_models(self) -> list[Model]:
@@ -1702,111 +1718,3 @@ class UpstreamProvider:
             Model object or None if not found
         """
         return self._models_by_id.get(model_id)
-
-
-class OpenAIUpstreamProvider(UpstreamProvider):
-    """Upstream provider specifically configured for OpenAI API."""
-
-    def __init__(self, api_key: str, provider_fee: float = 1.01):
-        self.upstream_name = "openai"
-        super().__init__(
-            base_url="https://api.openai.com/v1",
-            api_key=api_key,
-            provider_fee=provider_fee,
-        )
-
-    def transform_model_name(self, model_id: str) -> str:
-        """Strip 'openai/' prefix for OpenAI API compatibility."""
-        return model_id.removeprefix("openai/")
-
-    async def fetch_models(self) -> list[Model]:
-        """Fetch OpenAI models from OpenRouter API filtered by openai source."""
-        models_data = await async_fetch_openrouter_models(source_filter="openai")
-        return [Model(**model) for model in models_data]  # type: ignore
-
-
-class AnthropicUpstreamProvider(UpstreamProvider):
-    """Upstream provider specifically configured for Anthropic API."""
-
-    def __init__(self, api_key: str, provider_fee: float = 1.01):
-        self.upstream_name = "anthropic"
-        super().__init__(
-            base_url="https://api.anthropic.com/v1",
-            api_key=api_key,
-            provider_fee=provider_fee,
-        )
-
-    def transform_model_name(self, model_id: str) -> str:
-        """Strip 'anthropic/' prefix for Anthropic API compatibility."""
-        return model_id.removeprefix("anthropic/")
-
-    async def fetch_models(self) -> list[Model]:
-        """Fetch Anthropic models from OpenRouter API filtered by anthropic source."""
-        models_data = await async_fetch_openrouter_models(source_filter="anthropic")
-        return [Model(**model) for model in models_data]  # type: ignore
-
-
-class AzureUpstreamProvider(UpstreamProvider):
-    """Upstream provider specifically configured for Azure OpenAI Service."""
-
-    def __init__(
-        self,
-        base_url: str,
-        api_key: str,
-        api_version: str,
-        provider_fee: float = 1.01,
-    ):
-        """Initialize Azure provider with API key and version.
-
-        Args:
-            base_url: Azure OpenAI endpoint base URL
-            api_key: Azure OpenAI API key for authentication
-            api_version: Azure OpenAI API version (e.g., "2024-02-15-preview")
-            provider_fee: Provider fee multiplier (default 1.01 for 1% fee)
-        """
-        super().__init__(
-            base_url=base_url,
-            api_key=api_key,
-            provider_fee=provider_fee,
-        )
-        self.api_version = api_version
-
-    def prepare_params(
-        self, path: str, query_params: Mapping[str, str] | None
-    ) -> Mapping[str, str]:
-        """Prepare query parameters for Azure OpenAI, adding API version.
-
-        Args:
-            path: Request path
-            query_params: Original query parameters from the client
-
-        Returns:
-            Query parameters dict with Azure API version added for chat completions
-        """
-        params = dict(query_params or {})
-        if path.endswith("chat/completions"):
-            params["api-version"] = self.api_version
-        return params
-
-
-class OpenRouterUpstreamProvider(UpstreamProvider):
-    """Upstream provider specifically configured for OpenRouter API."""
-
-    def __init__(self, api_key: str, provider_fee: float = 1.06):
-        """Initialize OpenRouter provider with API key.
-
-        Args:
-            api_key: OpenRouter API key for authentication
-            provider_fee: Provider fee multiplier (default 1.06 for 6% fee)
-        """
-        self.upstream_name = "openrouter"
-        super().__init__(
-            base_url="https://openrouter.ai/api/v1",
-            api_key=api_key,
-            provider_fee=provider_fee,
-        )
-
-    async def fetch_models(self) -> list[Model]:
-        """Fetch all OpenRouter models."""
-        models_data = await async_fetch_openrouter_models()
-        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/fireworks.py

@@ -0,0 +1,24 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class FireworksUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for Fireworks.ai API."""
+
+    upstream_name = "fireworks"
+    base_url = "https://api.fireworks.ai/inference/v1"
+    platform_url = "https://app.fireworks.ai/settings/users/api-keys"
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        super().__init__(
+            base_url=self.base_url, api_key=api_key, provider_fee=provider_fee
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'fireworks/' prefix for Fireworks API compatibility."""
+        return model_id.removeprefix("fireworks/")
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch Fireworks models from OpenRouter API filtered by fireworks source."""
+        models_data = await async_fetch_openrouter_models(source_filter="fireworks")
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/generic.py

@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING
 
 import httpx
 
-from .upstream import UpstreamProvider
+from .base import BaseUpstreamProvider
 
 if TYPE_CHECKING:
     from ..payment.models import Model
@@ -14,7 +14,7 @@ from ..core.logging import get_logger
 logger = get_logger(__name__)
 
 
-class GenericUpstreamProvider(UpstreamProvider):
+class GenericUpstreamProvider(BaseUpstreamProvider):
     """Generic upstream provider that can fetch models from any OpenAI-compatible API."""
 
     def __init__(

routstr/upstream/groq.py

@@ -0,0 +1,24 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class GroqUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for Groq API."""
+
+    upstream_name = "groq"
+    base_url = "https://api.groq.com/openai/v1"
+    platform_url = "https://console.groq.com/keys"
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        super().__init__(
+            base_url=self.base_url, api_key=api_key, provider_fee=provider_fee
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'groq/' prefix for Groq API compatibility."""
+        return model_id.removeprefix("groq/")
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch Groq models from OpenRouter API filtered by groq source."""
+        models_data = await async_fetch_openrouter_models(source_filter="groq")
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/helpers.py

@@ -5,26 +5,26 @@ import re
 from typing import TYPE_CHECKING
 
 if TYPE_CHECKING:
-    from .core.settings import Settings
+    from ..core.settings import Settings
 
 
-from .core import get_logger
-from .core.db import AsyncSession, ModelRow, UpstreamProviderRow, create_session
-from .payment.models import Model
-from .upstreams import (
-    AnthropicUpstreamProvider,
-    AzureUpstreamProvider,
-    OllamaUpstreamProvider,
-    OpenAIUpstreamProvider,
-    OpenRouterUpstreamProvider,
-    UpstreamProvider,
-)
-from .upstreams.generic import GenericUpstreamProvider
+from ..core import get_logger
+from ..core.db import AsyncSession, ModelRow, UpstreamProviderRow, create_session
+from ..payment.models import Model
+from .anthropic import AnthropicUpstreamProvider
+from .azure import AzureUpstreamProvider
+from .base import BaseUpstreamProvider
+from .generic import GenericUpstreamProvider
+from .ollama import OllamaUpstreamProvider
+from .openai import OpenAIUpstreamProvider
+from .openrouter import OpenRouterUpstreamProvider
 
 logger = get_logger(__name__)
 
 
-def resolve_model_alias(model_id: str, canonical_slug: str | None = None) -> list[str]:
+def resolve_model_alias(
+    model_id: str, canonical_slug: str | None = None, alias_ids: list[str] | None = None
+) -> list[str]:
     """Resolve model ID to all possible aliases.
 
     Returns list of aliases including canonical slug and variations without provider prefix.
@@ -66,11 +66,14 @@ def resolve_model_alias(model_id: str, canonical_slug: str | None = None) -> lis
                 if canonical_base not in aliases:
                     aliases.append(canonical_base)
 
+    if alias_ids:
+        aliases.extend(alias_ids)
+
     return aliases
 
 
 async def get_all_models_with_overrides(
-    upstreams: list[UpstreamProvider],
+    upstreams: list[BaseUpstreamProvider],
 ) -> list[Model]:
     """Get all models from all providers with database overrides applied.
 
@@ -85,7 +88,7 @@ async def get_all_models_with_overrides(
     """
     from sqlmodel import select
 
-    from .payment.models import _row_to_model
+    from ..payment.models import _row_to_model
 
     async with create_session() as session:
         result = await session.exec(select(ModelRow).where(ModelRow.enabled))
@@ -120,57 +123,8 @@ async def get_all_models_with_overrides(
     return list(all_models.values())
 
 
-async def get_model_with_override(
-    model_id: str,
-    upstreams: list[UpstreamProvider],
-    session: AsyncSession,
-) -> Model | None:
-    """Get a specific model from providers with database override applied.
-
-    Resolves model aliases automatically (e.g., both "gpt-5-mini" and "openai/gpt-5-mini").
-
-    Args:
-        model_id: Model identifier (with or without provider prefix)
-        upstreams: List of upstream provider instances
-
-    Returns:
-        Model object or None if not found
-    """
-    from sqlmodel import select
-
-    from .payment.models import _row_to_model
-
-    aliases = resolve_model_alias(model_id)
-
-    for alias in aliases:
-        result = await session.exec(
-            select(ModelRow).where(
-                ModelRow.id == alias,
-                ModelRow.upstream_provider_id.isnot(None),  # type: ignore
-                ModelRow.enabled,
-            )
-        )
-        override_row = result.first()
-        if override_row:
-            provider = await session.get(
-                UpstreamProviderRow, override_row.upstream_provider_id
-            )
-            provider_fee = provider.provider_fee if provider else 1.01
-            return _row_to_model(
-                override_row, apply_provider_fee=True, provider_fee=provider_fee
-            )
-
-    for alias in aliases:
-        for upstream in upstreams:
-            model = upstream.get_cached_model_by_id(alias)
-            if model and model.enabled:
-                return model
-
-    return None
-
-
 async def refresh_upstreams_models_periodically(
-    upstreams: list[UpstreamProvider],
+    upstreams: list[BaseUpstreamProvider],
 ) -> None:
     """Background task to periodically refresh models cache for all providers.
 
@@ -180,7 +134,7 @@ async def refresh_upstreams_models_periodically(
     import asyncio
     import random
 
-    from .core.settings import settings
+    from ..core.settings import settings
 
     interval = getattr(settings, "models_refresh_interval_seconds", 0)
     if not interval or interval <= 0:
@@ -212,7 +166,7 @@ async def refresh_upstreams_models_periodically(
             break
 
 
-async def init_upstreams() -> list[UpstreamProvider]:
+async def init_upstreams() -> list[BaseUpstreamProvider]:
     """Initialize upstream providers from database.
 
     Seeds database with providers from settings if empty, then loads and instantiates
@@ -220,7 +174,7 @@ async def init_upstreams() -> list[UpstreamProvider]:
     """
     from sqlmodel import select
 
-    from .core.settings import settings
+    from ..core.settings import settings
 
     async with create_session() as session:
         result = await session.exec(select(UpstreamProviderRow))
@@ -235,7 +189,7 @@ async def init_upstreams() -> list[UpstreamProvider]:
             result = await session.exec(select(UpstreamProviderRow))
             existing_providers = result.all()
 
-        upstreams: list[UpstreamProvider] = []
+        upstreams: list[BaseUpstreamProvider] = []
         for provider_row in existing_providers:
             if not provider_row.enabled:
                 logger.debug(f"Skipping disabled provider: {provider_row.base_url}")
@@ -266,7 +220,7 @@ async def _seed_providers_from_settings(
     """
     from sqlmodel import select
 
-    from .core.settings import settings
+    from ..core.settings import settings
 
     providers_to_add: list[UpstreamProviderRow] = []
     seeded_base_urls: set[str] = set()
@@ -415,7 +369,9 @@ async def _seed_providers_from_settings(
         )
 
 
-def _instantiate_provider(provider_row: UpstreamProviderRow) -> UpstreamProvider | None:
+def _instantiate_provider(
+    provider_row: UpstreamProviderRow,
+) -> BaseUpstreamProvider | None:
     """Instantiate an UpstreamProvider from a database row.
 
     Args:
@@ -462,7 +418,7 @@ def _instantiate_provider(provider_row: UpstreamProviderRow) -> UpstreamProvider
                 provider_row.provider_type,
             )
         elif provider_row.provider_type == "custom":
-            return UpstreamProvider(
+            return BaseUpstreamProvider(
                 provider_row.base_url, provider_row.api_key, provider_row.provider_fee
             )
         else:

routstr/upstream/ollama.py

@@ -6,7 +6,7 @@ import httpx
 from fastapi import Request
 from fastapi.responses import Response, StreamingResponse
 
-from .upstream import UpstreamProvider
+from .base import BaseUpstreamProvider
 
 if TYPE_CHECKING:
     from ..core.db import ApiKey, AsyncSession
@@ -17,7 +17,7 @@ from ..core.logging import get_logger
 logger = get_logger(__name__)
 
 
-class OllamaUpstreamProvider(UpstreamProvider):
+class OllamaUpstreamProvider(BaseUpstreamProvider):
     """Upstream provider specifically configured for Ollama API."""
 
     def __init__(
@@ -121,7 +121,7 @@ class OllamaUpstreamProvider(UpstreamProvider):
                     models_list.append(
                         Model(
                             id=model_name,
-                            name=model_name,
+                            name=model_name.replace(":", " "),
                             created=0,
                             description=description,
                             context_length=context_length,

routstr/upstream/openai.py

@@ -0,0 +1,23 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class OpenAIUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for OpenAI API."""
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        self.upstream_name = "openai"
+        super().__init__(
+            base_url="https://api.openai.com/v1",
+            api_key=api_key,
+            provider_fee=provider_fee,
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'openai/' prefix for OpenAI API compatibility."""
+        return model_id.removeprefix("openai/")
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch OpenAI models from OpenRouter API filtered by openai source."""
+        models_data = await async_fetch_openrouter_models(source_filter="openai")
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/openrouter.py

@@ -0,0 +1,25 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class OpenRouterUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for OpenRouter API."""
+
+    def __init__(self, api_key: str, provider_fee: float = 1.06):
+        """Initialize OpenRouter provider with API key.
+
+        Args:
+            api_key: OpenRouter API key for authentication
+            provider_fee: Provider fee multiplier (default 1.06 for 6% fee)
+        """
+        self.upstream_name = "openrouter"
+        super().__init__(
+            base_url="https://openrouter.ai/api/v1",
+            api_key=api_key,
+            provider_fee=provider_fee,
+        )
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch all OpenRouter models."""
+        models_data = await async_fetch_openrouter_models()
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/perplexity.py

@@ -0,0 +1,26 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class PerplexityUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for OpenAI API."""
+
+    upstream_name = "perplexity"
+    base_url = "https://api.perplexity.ai/"  # without v1
+    platform_url = "https://www.perplexity.ai/account/api/keys"
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        super().__init__(
+            base_url=self.base_url,
+            api_key=api_key,
+            provider_fee=provider_fee,
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'perplexity/' prefix for Perplexity API compatibility."""
+        return model_id.removeprefix("perplexity/")
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch Perplexity models from OpenRouter API filtered by perplexity source."""
+        models_data = await async_fetch_openrouter_models(source_filter="perplexity")
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstream/xai.py

@@ -0,0 +1,24 @@
+from ..payment.models import Model, async_fetch_openrouter_models
+from .base import BaseUpstreamProvider
+
+
+class XAIUpstreamProvider(BaseUpstreamProvider):
+    """Upstream provider specifically configured for XAI API."""
+
+    upstream_name = "xai"
+    base_url = "https://api.x.ai/v1"
+    platform_url = "https://accounts.x.ai/sign-up"
+
+    def __init__(self, api_key: str, provider_fee: float = 1.01):
+        super().__init__(
+            base_url=self.base_url, api_key=api_key, provider_fee=provider_fee
+        )
+
+    def transform_model_name(self, model_id: str) -> str:
+        """Strip 'xai/' prefix for XAI API compatibility."""
+        return model_id.removeprefix("xai/")
+
+    async def fetch_models(self) -> list[Model]:
+        """Fetch XAI models from OpenRouter API filtered by xai source."""
+        models_data = await async_fetch_openrouter_models(source_filter="xai")
+        return [Model(**model) for model in models_data]  # type: ignore

routstr/upstreams/__init__.py

@@ -1,17 +0,0 @@
-from .ollama import OllamaUpstreamProvider
-from .upstream import (
-    AnthropicUpstreamProvider,
-    AzureUpstreamProvider,
-    OpenAIUpstreamProvider,
-    OpenRouterUpstreamProvider,
-    UpstreamProvider,
-)
-
-__all__ = [
-    "OllamaUpstreamProvider",
-    "UpstreamProvider",
-    "AnthropicUpstreamProvider",
-    "AzureUpstreamProvider",
-    "OpenAIUpstreamProvider",
-    "OpenRouterUpstreamProvider",
-]

scripts/build-ui.sh

@@ -49,6 +49,7 @@ else
     npm run build
 fi
 
+rm -rf ../ui_out
 mkdir -p ../ui_out
 mv out/* ../ui_out
 

ui/app/layout.tsx

@@ -34,7 +34,7 @@ export default function RootLayout({
   return (
     <html lang='en' suppressHydrationWarning>
       <body
-        className={`${geistSans.variable} ${geistMono.variable} antialiased font-sans`}
+        className={`${geistSans.variable} ${geistMono.variable} font-sans antialiased`}
       >
         <SuppressHydrationWarning>
           <Providers>{children}</Providers>

ui/app/model/page.tsx

@@ -16,6 +16,7 @@ import { Alert, AlertDescription } from '@/components/ui/alert';
 import { Badge } from '@/components/ui/badge';
 import { useMemo, useState } from 'react';
 import type { Model } from '@/lib/api/schemas/models';
+import { groupAndSortModelsByProvider } from '@/lib/utils/modelSort';
 
 export default function ModelsPage() {
   const [filteredModels, setFilteredModels] = useState<Model[]>([]);
@@ -34,15 +35,7 @@ export default function ModelsPage() {
 
   const groupedModels = useMemo(() => {
     if (!models) return {};
-
-    return models.reduce<Record<string, typeof models>>((acc, model) => {
-      const provider = model.provider;
-      if (!acc[provider]) {
-        acc[provider] = [];
-      }
-      acc[provider].push(model);
-      return acc;
-    }, {});
+    return groupAndSortModelsByProvider(models);
   }, [models]);
 
   const groupDataMap = useMemo(() => {
@@ -52,7 +45,9 @@ export default function ModelsPage() {
   const providerInfo = useMemo(() => {
     return Object.entries(groupedModels).map(([provider, providerModels]) => {
       const groupData = groupDataMap.get(provider);
-      const activeModels = providerModels.filter((m) => !m.soft_deleted).length;
+      const activeModels = providerModels.filter(
+        (m) => m.isEnabled && !m.soft_deleted
+      ).length;
       const totalModels = providerModels.length;
 
       return {
@@ -156,7 +151,10 @@ export default function ModelsPage() {
                             models={models}
                             onFilteredModelsChange={setFilteredModels}
                           />
-                          <ModelSelector filteredModels={filteredModels} />
+                          <ModelSelector
+                            filteredModels={filteredModels}
+                            showDeleteAllButton={true}
+                          />
                         </div>
                       </TabsContent>
 
@@ -183,7 +181,7 @@ export default function ModelsPage() {
                                               (m) => m.soft_deleted
                                             ).length
                                           }{' '}
-                                          soft deleted
+                                          disabled
                                         </span>
                                       )}
                                       {groupData?.group_url && (
@@ -199,6 +197,7 @@ export default function ModelsPage() {
                                   filterProvider={provider}
                                   groupData={groupData}
                                   showProviderActions={true}
+                                  showDeleteAllButton={false}
                                 />
                               </div>
                             </TabsContent>

ui/app/providers/page.tsx

@@ -277,7 +277,11 @@ export default function ProvidersPage() {
                         }
                         placeholder='https://api.example.com/v1'
                         disabled={hasFixedBaseUrl(formData.provider_type)}
-                        className={hasFixedBaseUrl(formData.provider_type) ? 'cursor-not-allowed opacity-60' : ''}
+                        className={
+                          hasFixedBaseUrl(formData.provider_type)
+                            ? 'cursor-not-allowed opacity-60'
+                            : ''
+                        }
                       />
                     </div>
                     <div className='grid gap-2'>
@@ -453,7 +457,8 @@ export default function ProvidersPage() {
                                 <div className='space-y-2'>
                                   {providerModels.db_models.length === 0 ? (
                                     <div className='text-muted-foreground py-4 text-center text-sm'>
-                                      No models configured. Add custom models to use this provider.
+                                      No models configured. Add custom models to
+                                      use this provider.
                                     </div>
                                   ) : (
                                     <div className='space-y-2'>
@@ -495,7 +500,10 @@ export default function ProvidersPage() {
                                 </div>
                               ) : (
                                 // Has provided models - show tabs
-                                <Tabs defaultValue='provided' className='w-full'>
+                                <Tabs
+                                  defaultValue='provided'
+                                  className='w-full'
+                                >
                                   <TabsList className='grid w-full grid-cols-2'>
                                     <TabsTrigger
                                       value='provided'
@@ -504,7 +512,9 @@ export default function ProvidersPage() {
                                       <span className='hidden sm:inline'>
                                         Provided Models
                                       </span>
-                                      <span className='sm:hidden'>Provided</span>
+                                      <span className='sm:hidden'>
+                                        Provided
+                                      </span>
                                       <Badge
                                         variant='secondary'
                                         className='ml-1 text-xs sm:ml-2'
@@ -534,7 +544,8 @@ export default function ProvidersPage() {
                                   >
                                     {providerModels.db_models.length > 0 && (
                                       <div className='text-muted-foreground mb-3 text-sm'>
-                                        Custom models override or extend the provider&apos;s catalog.
+                                        Custom models override or extend the
+                                        provider&apos;s catalog.
                                       </div>
                                     )}
                                     {providerModels.db_models.length === 0 ? (
@@ -543,39 +554,42 @@ export default function ProvidersPage() {
                                       </div>
                                     ) : (
                                       <div className='space-y-2'>
-                                        {providerModels.db_models.map((model) => (
-                                          <div
-                                            key={model.id}
-                                            className='hover:bg-accent flex flex-col gap-2 rounded-lg border p-3 transition-colors sm:flex-row sm:items-center sm:justify-between'
-                                          >
-                                            <div className='min-w-0 flex-1'>
-                                              <div className='flex flex-col gap-1 sm:flex-row sm:items-center sm:gap-2'>
-                                                <span className='truncate font-mono text-sm font-medium'>
-                                                  {model.id}
-                                                </span>
-                                                <Badge
-                                                  variant={
-                                                    model.enabled
-                                                      ? 'default'
-                                                      : 'secondary'
-                                                  }
-                                                  className='w-fit text-xs'
-                                                >
-                                                  {model.enabled
-                                                    ? 'Enabled'
-                                                    : 'Disabled'}
-                                                </Badge>
+                                        {providerModels.db_models.map(
+                                          (model) => (
+                                            <div
+                                              key={model.id}
+                                              className='hover:bg-accent flex flex-col gap-2 rounded-lg border p-3 transition-colors sm:flex-row sm:items-center sm:justify-between'
+                                            >
+                                              <div className='min-w-0 flex-1'>
+                                                <div className='flex flex-col gap-1 sm:flex-row sm:items-center sm:gap-2'>
+                                                  <span className='truncate font-mono text-sm font-medium'>
+                                                    {model.id}
+                                                  </span>
+                                                  <Badge
+                                                    variant={
+                                                      model.enabled
+                                                        ? 'default'
+                                                        : 'secondary'
+                                                    }
+                                                    className='w-fit text-xs'
+                                                  >
+                                                    {model.enabled
+                                                      ? 'Enabled'
+                                                      : 'Disabled'}
+                                                  </Badge>
+                                                </div>
+                                                <div className='text-muted-foreground mt-1 text-xs break-words'>
+                                                  {model.description ||
+                                                    model.name}
+                                                </div>
                                               </div>
-                                              <div className='text-muted-foreground mt-1 text-xs break-words'>
-                                                {model.description || model.name}
+                                              <div className='text-muted-foreground text-xs whitespace-nowrap'>
+                                                {model.context_length?.toLocaleString()}{' '}
+                                                tokens
                                               </div>
                                             </div>
-                                            <div className='text-muted-foreground text-xs whitespace-nowrap'>
-                                              {model.context_length?.toLocaleString()}{' '}
-                                              tokens
-                                            </div>
-                                          </div>
-                                        ))}
+                                          )
+                                        )}
                                       </div>
                                     )}
                                   </TabsContent>
@@ -583,9 +597,11 @@ export default function ProvidersPage() {
                                     value='provided'
                                     className='mt-4 space-y-2'
                                   >
-                                    {providerModels.remote_models.length > 0 && (
+                                    {providerModels.remote_models.length >
+                                      0 && (
                                       <div className='text-muted-foreground mb-3 text-sm'>
-                                        Models automatically discovered from the provider&apos;s catalog.
+                                        Models automatically discovered from the
+                                        provider&apos;s catalog.
                                       </div>
                                     )}
                                     <div className='space-y-2'>
@@ -670,7 +686,11 @@ export default function ProvidersPage() {
                   }
                   placeholder='https://api.example.com/v1'
                   disabled={hasFixedBaseUrl(formData.provider_type)}
-                  className={hasFixedBaseUrl(formData.provider_type) ? 'cursor-not-allowed opacity-60' : ''}
+                  className={
+                    hasFixedBaseUrl(formData.provider_type)
+                      ? 'cursor-not-allowed opacity-60'
+                      : ''
+                  }
                 />
               </div>
               <div className='grid gap-2'>

ui/components/ModelSelector.tsx

@@ -59,18 +59,24 @@ import {
 } from 'lucide-react';
 import { toast } from 'sonner';
 import { cn } from '@/lib/utils';
+import {
+  sortModels,
+  groupAndSortModelsByProvider,
+} from '@/lib/utils/modelSort';
 
 interface ModelSelectorProps {
   filterProvider?: string;
   groupData?: ModelGroup;
   showProviderActions?: boolean;
   filteredModels?: Model[];
+  showDeleteAllButton?: boolean;
 }
 
 export function ModelSelector({
   filterProvider,
   groupData,
   filteredModels: propFilteredModels,
+  showDeleteAllButton = false,
 }: ModelSelectorProps) {
   const [selectedModelId, setSelectedModelId] = useState<string>('');
   const [, setHoveredModelId] = useState<string | null>(null);
@@ -387,14 +393,10 @@ export function ModelSelector({
               providerIdNum,
               model.id
             );
-            await AdminService.updateProviderModel(
-              providerIdNum,
-              model.full_name,
-              {
-                ...existingModel,
-                enabled: true,
-              }
-            );
+            await AdminService.updateProviderModel(providerIdNum, model.id, {
+              ...existingModel,
+              enabled: true,
+            });
             totalEnabled++;
           } catch (error) {
             console.error(`Failed to enable model ${model.full_name}:`, error);
@@ -426,20 +428,13 @@ export function ModelSelector({
           : providerFilteredModels;
 
     if (filterProvider) {
-      // If filtering by provider, return single group
-      return { [filterProvider]: modelsToGroup };
+      const sortedModels = sortModels(modelsToGroup);
+      return { [filterProvider]: sortedModels };
     }
 
     if (!modelsToGroup) return {};
 
-    return modelsToGroup.reduce<Record<string, Model[]>>((acc, model) => {
-      const provider = model.provider;
-      if (!acc[provider]) {
-        acc[provider] = [];
-      }
-      acc[provider].push(model);
-      return acc;
-    }, {});
+    return groupAndSortModelsByProvider(modelsToGroup);
   }, [
     providerFilteredModels,
     filteredModels,
@@ -828,13 +823,15 @@ export function ModelSelector({
           <Square className='mr-2 h-4 w-4' />
           Deselect All
         </Button>
-        <Button
-          onClick={handleDeleteAll}
-          className='text-destructive focus:text-destructive'
-        >
-          <AlertTriangle className='mr-2 h-4 w-4' />
-          Delete All Models Permanently
-        </Button>
+        {showDeleteAllButton && (
+          <Button
+            onClick={handleDeleteAll}
+            className='text-destructive focus:text-destructive'
+          >
+            <AlertTriangle className='mr-2 h-4 w-4' />
+            Delete All Overrides Permanently
+          </Button>
+        )}
         {/* Model Management Actions
         <Button onClick={() => setIsAddFormOpen(true)} className='gap-2'>
           <Plus className='h-4 w-4' />
@@ -1077,9 +1074,9 @@ export function ModelSelector({
                                 variant='ghost'
                                 size='sm'
                                 onClick={(e) => e.stopPropagation()}
-                                className='h-8 w-8 p-0'
+                                className='hover:bg-muted/50 dark:hover:bg-muted/80 h-8 w-8 p-0'
                               >
-                                <MoreVertical className='h-4 w-4' />
+                                <MoreVertical className='text-muted-foreground hover:text-foreground h-4 w-4' />
                               </Button>
                             </DropdownMenuTrigger>
                             <DropdownMenuContent align='end'>
@@ -1181,7 +1178,7 @@ export function ModelSelector({
                         {model.soft_deleted && (
                           <span className='inline-flex items-center rounded-full border border-red-300 bg-red-100 px-2.5 py-0.5 text-xs font-medium text-red-800'>
                             <Trash2 className='mr-1 h-3 w-3' />
-                            Deleted
+                            Disabled
                           </span>
                         )}
                       </div>

ui/components/detailed-wallet-balance.tsx

@@ -203,7 +203,8 @@ export function DetailedWalletBalance({
                             <div
                               className={cn(
                                 'text-right font-mono',
-                                !detail.error && ownerMsat > 0 &&
+                                !detail.error &&
+                                  ownerMsat > 0 &&
                                   'font-semibold text-green-600'
                               )}
                             >
@@ -250,7 +251,8 @@ export function DetailedWalletBalance({
                                 <div
                                   className={cn(
                                     'truncate font-mono text-sm',
-                                    !detail.error && ownerMsat > 0 &&
+                                    !detail.error &&
+                                      ownerMsat > 0 &&
                                       'font-semibold text-green-600'
                                   )}
                                 >
@@ -283,4 +285,4 @@ export function DetailedWalletBalance({
       />
     </>
   );
-}
+}

ui/components/settings/admin-settings.tsx

@@ -288,22 +288,6 @@ export function AdminSettings() {
           </CardContent>
         </Card>
 
-        {/* Security Settings */}
-        <Card>
-          <CardHeader>
-            <CardTitle>Security Settings</CardTitle>
-            <CardDescription>
-              Configure authentication and API access
-            </CardDescription>
-          </CardHeader>
-          <CardContent className='space-y-4'>
-            {renderSecretField(
-              'upstream_api_key',
-              'Upstream API Key',
-              'Enter API key'
-            )}
-          </CardContent>
-        </Card>
         {/* Cashu Mints */}
         <Card>
           <CardHeader>

ui/components/temporary-balances.tsx

@@ -11,10 +11,7 @@ import {
   DollarSign,
   Activity,
 } from 'lucide-react';
-import {
-  AdminService,
-  TemporaryBalance,
-} from '@/lib/api/services/admin';
+import { AdminService, TemporaryBalance } from '@/lib/api/services/admin';
 import {
   Card,
   CardContent,
@@ -243,16 +240,16 @@ export function TemporaryBalances({
                             <div className='text-muted-foreground text-xs font-medium'>
                               Balance
                             </div>
-                          <div className='truncate font-mono text-sm'>
-                            {formatBalance(balance.balance)}
+                            <div className='truncate font-mono text-sm'>
+                              {formatBalance(balance.balance)}
                             </div>
                           </div>
                           <div className='space-y-1'>
                             <div className='text-muted-foreground text-xs font-medium'>
                               Spent
                             </div>
-                          <div className='truncate font-mono text-sm'>
-                            {formatBalance(balance.total_spent)}
+                            <div className='truncate font-mono text-sm'>
+                              {formatBalance(balance.total_spent)}
                             </div>
                           </div>
                         </div>

ui/lib/currency.ts

@@ -43,4 +43,3 @@ export function formatFromMsat(
   });
   return formatter.format(usd);
 }
-

ui/lib/exchange-rate.ts

@@ -51,4 +51,3 @@ export async function fetchBtcUsdPrice(): Promise<number | null> {
 export function btcToSatsRate(btcUsdPrice: number): number {
   return btcUsdPrice / 100_000_000;
 }
-

ui/lib/types/units.ts

@@ -14,4 +14,3 @@ export function getDisplayUnitLabel(unit: DisplayUnit): string {
       return unit;
   }
 }
-

ui/lib/utils/modelSort.ts

@@ -0,0 +1,36 @@
+import { type Model } from '@/lib/api/schemas/models';
+
+export function sortModelsByStatus(a: Model, b: Model): number {
+  if (a.isEnabled && !b.isEnabled) return -1;
+  if (!a.isEnabled && b.isEnabled) return 1;
+
+  if (a.isEnabled === b.isEnabled) {
+    if (!a.soft_deleted && b.soft_deleted) return -1;
+    if (a.soft_deleted && !b.soft_deleted) return 1;
+  }
+
+  return 0;
+}
+
+export function sortModels(models: Model[]): Model[] {
+  return [...models].sort(sortModelsByStatus);
+}
+
+export function groupAndSortModelsByProvider(
+  models: Model[]
+): Record<string, Model[]> {
+  const grouped = models.reduce<Record<string, Model[]>>((acc, model) => {
+    const provider = model.provider;
+    if (!acc[provider]) {
+      acc[provider] = [];
+    }
+    acc[provider].push(model);
+    return acc;
+  }, {});
+
+  Object.keys(grouped).forEach((provider) => {
+    grouped[provider].sort(sortModelsByStatus);
+  });
+
+  return grouped;
+}