experimental_decide | AI SDK 한국어

import { experimental_decide } from 'ai';

Decides answers to a nonempty map of choice, score, boolean, and box questions against one state. See Decisions for examples and semantics.

Parameters

Parameter	Type	Description
`model`	`Experimental_DecisionModel`	Required experimental v4 model instance or a string ID resolved by Gateway or an explicitly configured decision-capable default provider.
`state`	`Experimental_DecisionState`	Required string, JSON object, or ordered array of text, file, and JSON parts.
`questions`	`Record<string, Experimental_DecisionQuestion>`	Required nonempty question map.
`maxOutputTokens`	`number`	Optional maximum output tokens, an integer >= 1.
`reasoning`	`'provider-default' \| 'none' \| 'minimal' \| 'low' \| 'medium' \| 'high' \| 'xhigh' \| 'max'`	Optional reasoning effort. Language-model adapters default to `'none'`. Native decision providers return an unsupported-setting warning.
`maxRetries`	`number`	Nonnegative integer; defaults to 2.
`abortSignal`	`AbortSignal`	Cancels the decisions request.
`headers`	`Record<string, string>`	Additional HTTP headers.
`providerOptions`	`ProviderOptions`	Provider-specific options.
`telemetry`	`TelemetryOptions`	Telemetry configuration, including per-call integrations, input/output recording, and a function ID.
`runtimeContext`	`Record<string, unknown>`	Context available to lifecycle callbacks and selectively included in telemetry with `telemetry.includeRuntimeContext`.
`onStart`	`(event: Experimental_DecideStartEvent) => void`	Called when the decision operation begins.
`onEnd`	`(event: Experimental_DecideEndEvent) => void`	Called when the decision operation completes successfully.

Set reasoning: 'provider-default' to use the underlying language model's default reasoning effort instead of the adapter's 'none' default.

See Lifecycle Callbacks for the complete onStart and onEnd event fields.

Questions

Each question has a type and can include instructions and criteria as described below. Instructions and criterion descriptions accept strings, JSON objects, or JSON arrays.

Type	Instructions	Criteria
`choice`	Optional; may be `null`	Required nonempty map of option names to descriptions
`score`	Optional; may be `null`	Required nonempty array of levels, indexed from zero
`boolean`	Optional; may be `null`	Optional object with `true` and/or `false` descriptions; at least one must be non-null when instructions are absent
`box`	Required and non-null	None

Criterion descriptions may be null in the shared type, but providers can impose stricter limits. Perplexity requires non-null Score levels and accepts 1–10 levels; a single level yields score 0 and probability 1. Other providers may require at least two levels. Optional instructions also remain subject to the provider's API contract.

See question semantics and Perplexity decision examples.

Result

Returns Promise<Experimental_DecisionResult<QUESTIONS>>:

answers: One typed answer per question ID, with literal Choice option inference.
usage: inputTokens, outputTokens, and totalTokens, each possibly undefined.
warnings: Provider warnings, also passed to the SDK warning logger.
rounding: Optional provider-declared decimal precision for probabilities and scores.
providerMetadata: Optional provider-specific metadata.
response: Timestamp, model ID, and optional response ID, headers, and body.

Provider specification

Experimental_DecisionModelV4 is exported from @ai-sdk/provider and declares specificationVersion: 'v4', provider, modelId, supportedQuestionTypes, and doDecide(options). Decision models are isolated from stable ProviderV4.

The public core types are Experimental_DecisionModel, Experimental_DecisionState, Experimental_DecisionStatePart, Experimental_DecisionQuestion, Experimental_DecisionAnswer, and Experimental_DecisionResult. All decision-specific classes use the Decision prefix, with Experimental_ aliases at package boundaries.

Errors

Unsupported types throw Experimental_DecisionUnsupportedQuestionTypeError before provider I/O. Invalid inputs throw InvalidArgumentError; malformed answers throw InvalidResponseDataError. Questions the model refuses throw Experimental_DecisionRefusalError. Invalid answers and refusals are not retried. Neither partial results nor missing probability synthesis are supported.

Model resolution

Use registry.decisionModel('provider:model') or customProvider({ decisionModels: { alias: model } }).decisionModel('alias') to resolve models. Strings passed directly to experimental_decide use globalThis.AI_SDK_DEFAULT_PROVIDER.decisionModel(id) when a default provider is configured, or Gateway when no default provider is configured.

Resolution errors use the existing NoSuchModelError and NoSuchProviderError classes with modelType: 'decisionModel'. Model instances and resolved models must implement v4; other versions throw UnsupportedModelVersionError. See model resolution examples.