<br />

## Classes

| Class | Description |
|---|---|
| [DecisionMaker](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.decisionmaker) | MediaPipe Task that performs semantic decision making. |
| [FilesetResolver](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.filesetresolver) | Resolves the files required for the MediaPipe Task APIs. The returned filesets require that the Wasm files are published without renaming. If this is not possible, you can invoke the MediaPipe Tasks APIs using a manually created `WasmFileset`. |

## Interfaces

| Interface | Description |
|---|---|
| [BooleanQuestion](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.booleanquestion) | Question configuration for a Boolean Predicate evaluation. |
| [BooleanResult](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.booleanresult) | Result of evaluating a Boolean. |
| [ChoiceQuestion](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.choicequestion) | Question configuration for a Categorical Choice evaluation. |
| [ChoiceResult](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.choiceresult) | Result of evaluating a Choice. |
| [ClassifierDecision](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifierdecision) | Unified decision output for a single ClassifierQuestion. |
| [ClassifierEvaluateOptions](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifierevaluateoptions) | Per-call options for `evaluate()`, `evaluateBatch()`, and `classify()`. |
| [ClassifierExpectedInput](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifierexpectedinput) | Expected input modality and language configuration. |
| [ClassifierOption](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifieroption) | A single candidate option for binary, categorical, or ordinal questions. |
| [ClassifierOptionProbability](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifieroptionprobability) | Calibrated probability entry for a single candidate option. |
| [ClassifierQuestion](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifierquestion) | A single question within a multi-question ClassifierSchema. |
| [ClassifierSchema](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.classifierschema) | Multi-question decision schema compatible with the Web Classifier API. |
| [DecisionContextInput](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.decisioncontextinput) | Multimodal decision input combining text, encoded image bytes, and/or 16kHz mono PCM audio samples. |
| [DecisionMakerOptions](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.decisionmakeroptions) | Options to configure the MediaPipe Semantic Decision Maker Task. Note: On Web, `baseOptions.delegate` defaults to `'GPU'` (WebGPU). `.litertlm` containers with external weights (such as EmbeddingGemma 2 and Gemma 4) require WebGPU on Web and do not support `delegate: 'CPU'`. Setting `baseOptions.delegate` to `'CPU'` on Web is only supported for `.task` or inline `.tflite` models (such as Laya). |
| [ScoreQuestion](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.scorequestion) | Question configuration for an Ordinal Score evaluation. |
| [ScoreResult](https://developers.google.com/edge/api/mediapipe/js/tasks-decision.scoreresult) | Result of evaluating an Score. |

## Type Aliases

| Type Alias | Description |
|---|---|
| [ClassifierQuestionType](https://developers.google.com/edge/api/mediapipe/js/tasks-decision#classifierquestiontype) | Supported question types in a ClassifierSchema. |
| [ClassifierResult](https://developers.google.com/edge/api/mediapipe/js/tasks-decision#classifierresult) | Record mapping each question `id` in a `ClassifierSchema` to its `ClassifierDecision`. |

## ClassifierQuestionType {#classifierquestiontype}

Supported question types in a ClassifierSchema.

**Signature:**

    export declare type ClassifierQuestionType = "binary" | "categorical" | "ordinal" | "boolean" | "choice" | "score";

## ClassifierResult {#classifierresult}

Record mapping each question `id` in a `ClassifierSchema` to its `ClassifierDecision`.

**Signature:**

    export declare type ClassifierResult = Record<string, ClassifierDecision>;