Struct EmbeddingModel 

pub struct EmbeddingModel { /* private fields */ }

Text embedding model wrapper with configurable providers and LRU cache Supports hash-based embeddings (default), local ONNX models, and API services. Users can configure their preferred embedding method via manx config.

Implementations

impl EmbeddingModel

pub async fn new() -> Result<Self>

Create a new embedding model with default hash-based provider

pub async fn new_auto_select() -> Result<Self>

Create embedding model with smart auto-selection of best available provider

pub async fn new_with_config(config: EmbeddingConfig) -> Result<Self>

Create a new embedding model with custom configuration

pub async fn embed_text(&self, text: &str) -> Result<Vec<f32>>

Generate embeddings for a single text using configured provider with caching

pub async fn embed_batch(&self, texts: &[&str]) -> Result<Vec<Vec<f32>>>

Generate embeddings for multiple texts (batch processing) More efficient than calling embed_text repeatedly - uses native batch for ONNX providers

pub async fn get_dimension(&self) -> Result<usize>

Get the dimension of embeddings produced by this model

pub async fn health_check(&self) -> Result<()>

Test if the embedding model is working correctly

pub fn get_provider_info(&self) -> ProviderInfo

Get information about the current provider

pub fn get_config(&self) -> &EmbeddingConfig

Get the current configuration

pub fn cosine_similarity(a: &[f32], b: &[f32]) -> f32

Calculate cosine similarity between two embeddings

pub async fn auto_select_best_provider() -> Result<EmbeddingConfig>

Automatically select the best available embedding provider from installed models Respects user’s installed models and doesn’t hardcode specific model names