Module traits 

Core traits that every provider and model implementation must satisfy.

Re-exports

pub use raw_tensor_model::DimSize;

pub use raw_tensor_model::RawTensorModel;

pub use raw_tensor_model::TensorBatch;

pub use raw_tensor_model::TensorDtype;

pub use raw_tensor_model::TensorSpec;

pub use raw_tensor_model::TensorValue;

pub use asr::TranscribeOptions;

pub use asr::TranscribeResult;

pub use asr::TranscribeSegment;

pub use asr::TranscribeWord;

pub use asr::TranscriptionModel;

pub use docs::DocBlock;

pub use docs::DocBlockKind;

pub use docs::DocExtractOptions;

pub use docs::DocExtractResult;

pub use docs::DocOutputFormat;

pub use docs::DocumentExtractionModel;

pub use docs::OcrBlock;

pub use docs::OcrModel;

pub use docs::OcrResult;

pub use hybrid::HeadSet;

pub use hybrid::HybridEmbedResult;

pub use hybrid::HybridEmbeddingModel;

pub use multimodal::AudioEmbeddingModel;

pub use multimodal::AudioInput;

pub use multimodal::ImageEmbeddingModel;

pub use multimodal::Modality;

pub use multimodal::MultimodalBlock;

pub use multimodal::MultimodalEmbeddingModel;

pub use multimodal::MultimodalInput;

pub use multivector::MultiVectorEmbedResult;

pub use multivector::MultiVectorEmbeddingModel;

pub use nlp::DepLink;

pub use nlp::NerEntity;

pub use nlp::NlpLabelMaps;

pub use nlp::NlpModel;

pub use nlp::NlpRequest;

pub use nlp::NlpResult;

pub use nlp::NlpSentence;

pub use nlp::NlpTasks;

pub use nlp::NlpToken;

pub use nlp::SpeechAct;

pub use nlp::SrlFrame;

pub use nlp::SrlRole;

pub use sparse::SparseEmbedResult;

pub use sparse::SparseEmbeddingModel;

pub use sparse::SparseVector;

Modules

asr

Automatic speech recognition types and trait.

docs

Document understanding types and traits — VLM-based extraction and OCR.

hybrid

Single-pass hybrid embedding: the HybridEmbeddingModel trait and its HeadSet / HybridEmbedResult types.

multimodal

Multimodal embedding types and traits (image / audio / mixed-modality).

multivector

Multi-vector (late-interaction / ColBERT) embedding types and the MultiVectorEmbeddingModel trait.

nlp

Structured natural-language analysis types and traits.

raw_tensor_model

sparse

Learned-sparse text embedding types and the SparseEmbeddingModel trait.

Structs

AudioOutput

Audio output produced by a speech model.

EmbedResult

Result of a dense-embedding call, carrying the vectors plus optional usage.

GeneratedImage

An image produced by a generation call (e.g. from a diffusion model).

GenerationOptions

Sampling and length parameters for text generation.

GenerationResult

The output of a generation call.

Message

A single message in a conversation, containing one or more content blocks.

ProviderCapabilities

Advertised capabilities of a ModelProvider.

ScoredDoc

A single scored document returned by a RerankerModel.

TokenUsage

Token counts for a generation request.

Enums

ContentBlock

A single block of content within a Message.

ImageInput

Image data that can be passed as part of a ContentBlock.

MessageRole

The role of a message in a conversation.

ProviderHealth

Health status reported by a provider.

Traits

EmbeddingModel

A model that produces dense vector embeddings from text.

GeneratorModel

A model that generates text, images, or audio from a conversational message history.

ModelInfo

Metadata common to every loaded model handle.

ModelProvider

A pluggable backend that knows how to load models for one or more ModelTask types.

RerankerModel

A model that re-scores documents against a query for relevance ranking.

Type Aliases

LoadedModelHandle

A type-erased, reference-counted handle to a loaded model instance.