class · iOS 26.0+ · SDK 27.1
final class SystemLanguageModelAn on-device Apple Foundation Model capable of text generation tasks. The `SystemLanguageModel` refers to the on-device text foundation model that powers Apple Intelligence. Use ``default`` to access the base version of the model and perform general-purpose text generation tasks. To access a specialized version of the model, initialize the model with ``UseCase`` to perform tasks like ``UseCase/contentTagging``. Apple periodically updates `SystemLanguageModel` in routine OS updates to improve the on-device model's abilities and performance. Currently, there are 3 model versions that align with: - iOS, iPadOS, macOS, and visionOS **26.0 - 26.3** - iOS, iPadOS, macOS, visionOS **26.4** - iOS, iPadOS, macOS, and visionOS **27.0** For more information about how model versions affect your app, see <doc:updating-prompts-for-new-model-versions>. Before you use the model, you need to verify its availability. Model availability depends on whether the device and region supports Apple Intelligence. For a list of supported devices, see [Apple Intelligence](https://www.apple.com/apple-intelligence/). Use ``Availability`` to change what your app shows to people based on the availability condition: ```swift struct GenerativeView: View { // Create a reference to the system language model. private var model = SystemLanguageModel.default var body: some View { switch model.availability { case .available: // Show your intelligence UI. case .unavailable(.deviceNotEligible): // Show an alternative UI. case .unavailable(.modelNotReady): // The model isn't ready because it's downloading or because // of other system reasons. case .unavailable(let other): // The model is unavailable for an unknown reason. } } } ```
tokenCountFor(instructions: Instructions): Promise<number>nonisolated(nonsending) final func tokenCount(for instructions: Instructions) async throws -> IntReturns the token count for the specified instructions. - Parameter instructions: The instructions to calculate the token count for. - Returns: The token count for the instructions.
tokenCountFor79776(schema: GenerationSchema): Promise<number>nonisolated(nonsending) final func tokenCount(for schema: GenerationSchema) async throws -> IntReturns the token count for the specified schema. - Parameter schema: The schema to calculate the token count for. - Returns: The token count for the schema.
tokenCountFor31777(tools: Array<SDKHandle>): Promise<number>nonisolated(nonsending) final func tokenCount(for tools: [any Tool]) async throws -> IntReturns the token count for the specified tools. - Parameter tools: An array of tools to calculate the token count for. - Returns: The token count for the tools.
tokenCountFor15824(prompt: SDKHandle): Promise<number>nonisolated(nonsending) final func tokenCount(for prompt: some PromptRepresentable) async throws -> IntReturns the token count for the specified prompt. - Parameter prompt: The prompt to calculate the token count for. - Returns: The token count for the prompt.
get contextSize(): number@backDeployed(before: iOS 26.4, macOS 26.4, visionOS 26.4)
final var contextSize: Int { get }The maximum context size in tokens that the model supports. The context size represents the total number of tokens that can be used in a single session, including both input prompts and generated responses. - Returns: The context size, in tokens. - Throws: An error if the context size cannot be determined. Typically this is due to the model not being available or Apple Intelligence is disabled.
get isAvailable(): booleanfinal var isAvailable: Bool { get }A Boolean value that indicates whether the system is entirely ready.
get availability(): SystemLanguageModel_Availabilityfinal var availability: SystemLanguageModel.Availability { get }The availability of the language model.
get capabilities(): LanguageModelCapabilitiesfinal var capabilities: LanguageModelCapabilities { get }The capabilities of this language model. If a developer attempts to use capabilities that your model does not support, the system automatically throws an error for you instead of calling a respond method, like ``LanguageModelSession/respond(to:options:)-(Prompt,_)`` or ``LanguageModelSession/streamResponse(to:options:)-(Prompt,_)``.
supportsLocale(locale?: _Foundation_Locale): booleanfinal func supportsLocale(_ locale: Locale = Locale.current) -> BoolReturns a Boolean value that indicates whether the given locale is supported by the model. Use this method over ``supportedLanguages`` to check whether the given locale qualifies a person for using this model, as this method also takes language fallbacks into consideration.
get supportedLanguages(): Array<_Foundation_Locale_Language>final var supportedLanguages: Set<Locale.Language> { get }Languages that the model supports. To check if a given locale is considered supported by the model, use ``supportsLocale(_:)``, which also takes language fallbacks into consideration.
get executorConfiguration(): SystemLanguageModel_Executor_Configurationfinal var executorConfiguration: SystemLanguageModel.Executor.Configuration { get }A configuration for an executor capable of running this model.
static get default(): SystemLanguageModelstatic var `default`: SystemLanguageModel { get }The base version of the model. The base model is a generic model that is useful for a wide variety of applications, but is not specialized to any particular use case.
static init(useCase?: SystemLanguageModel_UseCase, guardrails?: SystemLanguageModel_Guardrails): SystemLanguageModelconvenience init(useCase: SystemLanguageModel.UseCase = .general, guardrails: SystemLanguageModel.Guardrails = Guardrails.default)Creates a system language model instance for a specific use case.
get variant(): SystemLanguageModel_Variantfinal var variant: SystemLanguageModel.Variant { get }The variant of the on-device model backing this instance.
1 declarations were excluded from these bindings. Each reason comes from the generator or the iOS SDK compiler.
SystemLanguageModel.tokenCount(for:)
This parameter hides a concrete Swift type.