configs
Classes
ImportConfigForm
Bases: BaseModel
Form for importing system configuration.
Attributes
config
The configuration dictionary to import. This should match the structure returned by the export endpoint.
Dict Fields
This dictionary contains the complete Open WebUI configuration that can be exported and imported. It includes all settings from the backend config system. For a complete reference of all possible keys and their descriptions, see the backend configuration in: owui_client/refs/owui_source_main/backend/open_webui/main.py
The config includes but is not limited to: - API integrations (Ollama, OpenAI, etc.) - Feature flags (folders, channels, notes, etc.) - Authentication settings (OAuth, LDAP, etc.) - RAG and retrieval configurations - Image generation and editing settings - Audio processing configurations - Web search and loader configurations - Code execution and interpreter settings - UI and permission settings
ConnectionsConfigForm
Bases: BaseModel
Configuration for system connections.
- Code Reference routers configs Classes ConfigsClient Functions
- Code Reference routers configs Classes ConfigsClient Functions set_connections_config
Attributes
ENABLE_DIRECT_CONNECTIONS
Whether to allow users to connect to their own OpenAI compatible API endpoints directly.
OAuthClientRegistrationForm
Bases: BaseModel
Form for registering an OAuth client.
- Code Reference routers configs Classes ConfigsClient Functions register_oauth_client
ToolServerConnection
Bases: BaseModel
Configuration for a single tool server connection.
- Code Reference models configs Classes ToolServersConfigForm Attributes TOOL_SERVER_CONNECTIONS
- Code Reference routers configs Classes ConfigsClient Functions verify_tool_servers_config
Attributes
auth_type
Authentication type. Common values: 'bearer', 'session', 'system_oauth', 'oauth_2.1'.
headers
Custom headers to send with requests to the tool server.
Dict Fields
Authorization(str, optional): Authorization header for tool server authenticationContent-Type(str, optional): Content type header for requestsX-OpenWebUI-User-Name(str, optional): User name header for user contextX-OpenWebUI-Chat-Id(str, optional): Chat ID header for context trackingAccept(str, optional): Accept header for response content type- Any other custom headers needed for specific tool server requirements
The headers dictionary allows customization of HTTP headers sent to tool servers. Common use cases include: - Adding authentication headers beyond the standard bearer token - Setting custom content types for specific API requirements - Forwarding user information headers when ENABLE_FORWARD_USER_INFO_HEADERS is enabled - Including chat context headers for tracking and logging purposes
When provided as a string, it should be a JSON-encoded dictionary.
config
Additional configuration for the connection.
Dict Fields
enable(bool, optional): Whether the tool server connection is enabled. Defaults to True.function_name_filter_list(str, optional): Comma-separated list of function names to filter/allow for this tool server. Used to restrict which functions from the tool server are exposed.access_control(dict, optional): Access control configuration for the tool server connection. Defines permissions and restrictions for user access.oauth_server_key(str, optional): OAuth server key for OAuth 2.1 authentication with MCP tool servers. Used during dynamic client registration.
The config dictionary provides additional connection-specific settings that control behavior, security, and functionality of the tool server integration.
ToolServersConfigForm
Bases: BaseModel
Configuration for tool servers.
- Code Reference routers configs Classes ConfigsClient Functions
- Code Reference routers configs Classes ConfigsClient Functions set_tool_servers_config
Attributes
TOOL_SERVER_CONNECTIONS
TOOL_SERVER_CONNECTIONS: List[ToolServerConnection]
List of configured tool server connections.
CodeInterpreterConfigForm
Bases: BaseModel
Configuration for code execution and interpreter.
- Code Reference routers configs Classes ConfigsClient Functions
- Code Reference routers configs Classes ConfigsClient Functions set_code_execution_config
Attributes
CODE_EXECUTION_ENGINE
Engine for code execution. Supported: 'pyodide', 'jupyter'.
CODE_EXECUTION_JUPYTER_URL
URL for Jupyter server (if engine is jupyter).
CODE_EXECUTION_JUPYTER_AUTH
Auth method for Jupyter. Supported: 'token', 'password', or empty/None.
CODE_EXECUTION_JUPYTER_AUTH_TOKEN
Token for Jupyter auth.
CODE_EXECUTION_JUPYTER_AUTH_PASSWORD
Password for Jupyter auth.
CODE_EXECUTION_JUPYTER_TIMEOUT
Timeout for code execution in seconds.
ENABLE_CODE_INTERPRETER
Enable code interpreter feature (e.g. for chat).
CODE_INTERPRETER_ENGINE
Engine for code interpreter. Supported: 'pyodide', 'jupyter'.
CODE_INTERPRETER_PROMPT_TEMPLATE
Custom prompt template for the code interpreter.
CODE_INTERPRETER_JUPYTER_URL
URL for Jupyter server (if interpreter engine is jupyter).
CODE_INTERPRETER_JUPYTER_AUTH
Auth method for Jupyter interpreter. Supported: 'token', 'password', or empty/None.
CODE_INTERPRETER_JUPYTER_AUTH_TOKEN
Token for Jupyter interpreter auth.
CODE_INTERPRETER_JUPYTER_AUTH_PASSWORD
Password for Jupyter interpreter auth.
ModelsConfigForm
Bases: BaseModel
Configuration for model defaults and ordering.
- Code Reference routers configs Classes ConfigsClient Functions
- Code Reference routers configs Classes ConfigsClient Functions set_models_config
Attributes
DEFAULT_MODELS
Comma-separated list of default model IDs (e.g. for new chats).
DEFAULT_PINNED_MODELS
Comma-separated list of pinned model IDs.
PromptSuggestion
Bases: BaseModel
A prompt suggestion for the chat interface.
- Code Reference
SetDefaultSuggestionsForm
Bases: BaseModel
Form for setting default prompt suggestions.
- Code Reference routers configs Classes ConfigsClient Functions set_default_suggestions
Attributes
BannerModel
Bases: BaseModel
Model representing a banner notification.
- Code Reference
Attributes
SetBannersForm
Bases: BaseModel
Form for setting banners.
- Code Reference routers configs Classes ConfigsClient Functions set_banners