Struct OpenAIEmbeddingClient 

pub struct OpenAIEmbeddingClient { /* private fields */ }

OpenAI embedding provider.

Uses text-embedding-3-small (1536-dim) by default. Swap model to text-embedding-3-large (3072-dim) for higher accuracy.

api_key is not exposed via Debug to prevent accidental logging.

Implementations

impl OpenAIEmbeddingClient

pub fn new_small(api_key: impl Into<String>) -> Self

Create with text-embedding-3-small (1536 dimensions).

pub fn new_large(api_key: impl Into<String>) -> Self

Create with text-embedding-3-large (3072 dimensions).

pub fn new_with_model( api_key: impl Into<String>, model: impl Into<String>, dim: usize, ) -> Self

Create with an explicit model name and embedding dimension.

Use this for legacy models (text-embedding-ada-002), reduced-dimension variants, or any future model not covered by new_small and new_large.

Panics

Panics if dim is zero.

pub fn from_env() -> Result<Self>

Create from environment variable OPENAI_API_KEY.

Always uses text-embedding-3-small (1536-dim). For a different model use new_with_model after reading the key manually.

Trait Implementations

impl EmbeddingProvider for OpenAIEmbeddingClient

fn dim(&self) -> usize

The dimensionality of the embedding vectors.

fn name(&self) -> &str

Provider name for display.

fn embed(&self, text: &str) -> Result<EmbeddingResult>

Embed a single text string.

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

Embed multiple texts in batch.