Options for the useStructuredCompletion hook.
API
interface UseStructuredCompletionOptions<Input, Schema extends s.SchemaOutput> {debounceTime?: number;debugName?: string;input: Input | null | undefined;model: ModelInput; retries?: number;schema: Schema;structuredOutput?: Chat.Api.StructuredOutputOptions;system: string;threadId?: string | undefined;tools?: Chat.AnyTool[];transport?: TransportOrFactory;ui?: boolean;}UseStructuredCompletionOptions
@typeInput@typeSchemas.SchemaOutputdebounceTime
numberThe debounce time between sends to the endpoint. default: 150
debugName
stringThe name of the hook, useful for debugging.
input
Input | null | undefinedThe input string to predict from.
model
ModelInputThe LLM model to use for the chat.
retries
numberNumber of retries if an error is received. default: 0
schema
SchemaThe schema to use for the chat.
structuredOutput
Chat.Api.StructuredOutputOptionsControls how the provider is asked to produce structured output.
system
stringThe system message to use for the chat.
threadId
string | undefinedOptional opaque AG-UI thread identity.
tools
Chat.AnyTool[]The tools to make available for the chat. default: []
transport
TransportOrFactoryOptional transport override for this hook.
ui
booleanWhether this completion should be treated as UI-generating.