ConduitLLM WebUI is a .NET Blazor application that provides a comprehensive web interface for configuring and managing all aspects of the ConduitLLM system. This interface allows administrators to configure providers, manage model mappings, monitor usage, and test LLM interactions—all without writing any code.
- URL:
http://localhost:5000(default) - Authentication: Based on configured security options
When accessing the WebUI for the first time:
- You'll be prompted to configure a master key if none exists
- You'll see a guided setup experience to configure your first provider
- The dashboard will display a clear call to action if no providers are configured
The WebUI features a consistent navigation layout:
- Dashboard: Home page with system overview and statistics
- Configuration: Manage providers, models, and router settings
- Virtual Keys: Create and manage API access keys
- Chat: Test LLM interactions directly
- About: Information about the ConduitLLM project
- User Profile: Access user-specific settings (if authentication is enabled)
- Notifications: View system notifications and alerts
- Settings: Access global application settings
The home page provides an overview of your ConduitLLM system:
- Status Summary: Quick view of system status
- Provider Statistics: Number of configured providers and their status
- Model Mapping Count: Total number of configured model mappings
- Recent Activity: Latest system events and API calls
- Usage Metrics: Charts showing recent usage patterns
Learn more about dashboard features
The central hub for system configuration with multiple tabs:
Manage LLM provider connections:
- List all configured providers
- Add new providers with API keys
- Edit existing provider settings
- Toggle provider active status
- Access provider documentation and API key links
Configure generic-to-specific model mappings:
- Create abstract model names (e.g., "chat-large")
- Map these to provider-specific models
- Configure which providers handle which models
- Manage model aliases and categories
Configure the LLM router:
- Select routing strategy (simple, random, round-robin)
- Configure model deployments
- Set up fallback paths between models
- Adjust routing weights
- View router health status
Configure system-wide settings:
- Master key management
- Default timeout values
- Logging configuration
- Cache settings
- Security options
Manage API access through virtual keys:
- View all virtual keys with filtering and sorting
- Check key status (active/inactive)
- Monitor budget usage
- See expiration status
- Create new virtual keys with custom settings
- Edit existing keys (name, budget, expiration)
- Reset key spending
- Revoke/delete keys
- Visualize key usage over time
- Track budget consumption
- Identify usage patterns
- Export usage data
Test LLM interactions directly:
- Select models to interact with
- Configure completion parameters (temperature, tokens)
- Send messages and view responses
- Compare outputs between models
- Save and load conversations
- Navigate to Configuration > Providers
- Click Add Provider
- Select the provider type (e.g., OpenAI, Anthropic)
- Enter provider name (for your reference)
- Enter API key (links are provided to get API keys)
- Optionally configure custom endpoint URL
- Click Save
- Navigate to Configuration > Model Mappings
- Click Add Mapping
- Enter a generic model name (e.g., "chat-large")
- Select the provider from the dropdown
- Select the provider-specific model (e.g., "gpt-4" for OpenAI)
- Set the mapping as active
- Click Save
- Navigate to Virtual Keys
- Click Create Key
- Enter a name for the key
- Optionally set a budget limit
- Choose a budget reset period (daily/monthly)
- Set an expiration date if desired
- Click Create
- Copy the generated key for API use
- Navigate to Chat
- Select the model to test
- Configure parameters (temperature, max tokens)
- Enter a system message (optional)
- Type your message in the chat input
- Click Send or press Enter
- View the model's response
- Continue the conversation as needed
Screenshot pending: This is a placeholder. Please add a real screenshot of the Provider Edit Form.
This form includes:
- Provider type selection
- Name field
- API key field (with copy/paste support)
- Endpoint URL field
- Helpful links for API documentation and key generation
- Test connection button
Screenshot pending: This is a placeholder. Please add a real screenshot of the Model Mapping Manager.
Features:
- Generic model name field
- Provider selection
- Provider-specific model selection
- Active status toggle
- Bulk operations for multiple mappings
Screenshot pending: This is a placeholder. Please add a real screenshot of the Virtual Key Creator.
Includes:
- Name field
- Budget amount input
- Budget period selection
- Expiration date picker
- Active status toggle
- Generated key display (after creation)
Screenshot pending: This is a placeholder. Please add a real screenshot of the Chat Interface.
Provides:
- Model selection dropdown
- Parameter configuration panel
- System message input
- Conversation history display
- Message input
- Response streaming (when supported)
The WebUI includes a notification system that alerts you to important events:
- System Events: Server starts, updates, configuration changes
- Error Alerts: API failures, configuration issues
- Budget Alerts: Virtual keys approaching or exceeding budgets
- Expiration Notices: Virtual keys nearing expiration
- Usage Milestones: Significant usage thresholds
Notifications appear in:
- The notification panel (accessed via the bell icon)
- Toast messages for critical alerts
- Dashboard summaries
- Email (if configured)
The WebUI supports several customization options:
- Light and dark mode toggle
- Color scheme customization
- Layout density options
- Table view customization
- Chart type selection
- Dashboard widget arrangement
- Metric display options
- Default models for chat
- Parameter presets
- Notification preferences
The WebUI can be configured with different authentication methods:
- Local authentication (coming soon – see issue #17)
- OAuth/OpenID Connect (coming soon – see issue #17)
- LDAP/Active Directory (coming soon – see issue #17)
- Single Sign-On integration (coming soon – see issue #17)
- Current: Master key authentication for admin access
Only master key authentication is currently available. Additional authentication methods are planned and tracked in issue #17.
Role-based access control can restrict access to sensitive operations:
- Admin role: Full access to all features
- Operator role: Can manage but not create providers and keys
- User role: Can use chat and view own usage
- Read-only role: Can view but not modify configuration
- API keys are masked in the UI (displayed as ******)
- Caching is disabled for pages with sensitive information
- Master key changes require confirmation
- Page Doesn't Load: Check server status and browser console
- Changes Not Saved: Verify form validation and server connectivity
- Provider Test Fails: Check API key and endpoint URL
- Chat Not Responding: Verify provider configuration and connectivity
The WebUI works best with modern browsers:
- Chrome/Edge (recommended)
- Firefox
- Safari
UI errors are logged to:
- Browser console (client-side issues)
- Server logs (API and backend issues)
- Network tab (API request problems)
- Access WebUI only on secure networks
- Change the master key periodically
- Use strong passwords for authentication
- Log out when not actively using the interface
- Clear browser cache periodically
- Don't leave the WebUI open for extended periods
- Limit the number of records displayed in tables
- Use filtering to narrow down large data sets
- Make configuration changes systematically
- Test changes in the Chat interface
- Document custom configurations
- Regularly review provider settings