Implementation:Explodinggradients Ragas BaseRagasEmbedding Class

Knowledge Sources	Explodinggradients_Ragas
Domains	Embeddings, NLP
Last Updated	2026-02-10 00:00 GMT

Overview

Abstract base class hierarchy and factory functions for embedding providers, supporting both modern and legacy (LangChain-compatible) interfaces.

Description

This module provides BaseRagasEmbedding (modern ABC for embedding providers with sync/async support), BaseRagasEmbeddings (legacy LangChain-compatible ABC), LangchainEmbeddingsWrapper (deprecated wrapper), HuggingfaceEmbeddings (sentence-transformer based), LlamaIndexEmbeddingsWrapper, and the embedding_factory function for unified creation across providers.

Usage

Use embedding_factory to create embedding instances by provider name, or extend BaseRagasEmbedding to implement custom embedding providers.

Code Reference

Source Location

• Repository: Explodinggradients_Ragas
• File: src/ragas/embeddings/base.py
• Lines: 29-874

Signature

class BaseRagasEmbedding(ABC):
    @abstractmethod
    def embed_text(self, text: str, **kwargs) -> List[float]:
        ...
    @abstractmethod
    async def aembed_text(self, text: str, **kwargs) -> List[float]:
        ...
    def embed_texts(self, texts: List[str], **kwargs) -> List[List[float]]:
        ...
    async def aembed_texts(self, texts: List[str], **kwargs) -> List[List[float]]:
        ...

def embedding_factory(
    provider: str = "openai",
    model: Optional[str] = None,
    run_config: Optional[RunConfig] = None,
    client: Optional[Any] = None,
    interface: str = "auto",
    cache: Optional[CacheInterface] = None,
    **kwargs,
) -> Union[BaseRagasEmbeddings, BaseRagasEmbedding]:
    ...

Import

from ragas.embeddings.base import BaseRagasEmbedding, embedding_factory
from ragas.embeddings import LangchainEmbeddingsWrapper, HuggingfaceEmbeddings

I/O Contract

Inputs

Name	Type	Required	Description
text	str	Yes	Single text to embed
texts	List[str]	Yes (for batch)	Multiple texts to embed
provider	str	No	Provider name (default "openai")
model	Optional[str]	No	Model name override
client	Optional[Any]	No	Pre-configured client

Outputs

Name	Type	Description
embed_text returns	List[float]	Embedding vector for single text
embed_texts returns	List[List[float]]	Embedding vectors for batch
embedding_factory returns	BaseRagasEmbedding	Configured embedding instance

Usage Examples

from ragas.embeddings.base import embedding_factory

# Create OpenAI embeddings
embeddings = embedding_factory(provider="openai", model="text-embedding-3-small")

# Embed a single text
vector = embeddings.embed_text("What is machine learning?")
print(len(vector))  # embedding dimension

# Batch embed
vectors = embeddings.embed_texts(["Hello", "World"])
print(len(vectors))  # 2