DecisionModels class abstract final

Verified official models accepted by DecisionMakerOptions.

Google's cross-encoder decision models, Laya and GLiNER2.5-Decide, and the EmbeddingGemma 2 bi-encoder, as Google's Decision Maker guide lists them. The s256 and s512 variants read up to 256 or 512 tokens.

Properties

hashCode → int

The hash code for this object.

no setter, inherited

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

Methods

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

toString() → String

A string representation of this object.

inherited

Operators

operator ==(Object other) → bool

The equality operator.

inherited

Constants

byName → const Map<String, DownloadAsset>

Every model above by the name an app lists to bundle it, under hooks.user_defines.mediapipe_decision.models in pubspec.yaml.

embeddingGemma2Text → const DownloadAsset

EmbeddingGemma 2, text, 270M parameters: 165 MB. A bi-encoder backend in Google's Decision Maker guide, and the model Google's web demo starts with: as fast as Laya here at a quarter of the download. From Google's LiteRT community on Hugging Face, Apache 2.0, pinned to its revision.

embeddingGemma2TextVision → const DownloadAsset

EmbeddingGemma 2, text and images, 440M parameters: 388 MB. The same text encoder as embeddingGemma2Text with a vision encoder Decision Maker never runs, so it answers as that model does, at more than twice the download. The same pin as RetrievalModels.embeddingGemma2TextVision, downloaded once for both packages.

glinerS256 → const DownloadAsset

GLiNER2.5-Decide, float16, 256 tokens: 981 MB.

glinerS512 → const DownloadAsset

GLiNER2.5-Decide, float16, 512 tokens: 1.08 GB.

layaS256 → const DownloadAsset

Laya, float32, 256 tokens: 678 MB. The model the gallery runs.

layaS512 → const DownloadAsset

Laya, float32, 512 tokens: 679 MB.