> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# embedding • AI Agent SDK

> embedding: Generate embeddings for text using LiteLLM.

# embedding

<div className="flex items-center gap-2">
  <Badge color="teal">Function</Badge>
</div>

> This function is defined in the [**embed**](../modules/embed) module.

Generate embeddings for text using LiteLLM.

This is the primary embedding function that supports all LiteLLM
embedding providers (OpenAI, Azure, Cohere, Voyage, etc.).

## Signature

```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
def embedding(input: Union[str, List[str]], model: str, dimensions: Optional[int], encoding_format: str, timeout: float, api_key: Optional[str], api_base: Optional[str], metadata: Optional[Dict[str, Any]]) -> EmbeddingResult
```

## Parameters

<ParamField query="input" type="Union" required={true}>
  Text or list of texts to embed
</ParamField>

<ParamField query="model" type="str" required={false} default="'text-embedding-3-small'">
  Model name (e.g., "text-embedding-3-small", "text-embedding-3-large")
</ParamField>

<ParamField query="dimensions" type="Optional" required={false}>
  Optional output dimensions (for models that support it)
</ParamField>

<ParamField query="encoding_format" type="str" required={false} default="'float'">
  "float" or "base64"
</ParamField>

<ParamField query="timeout" type="float" required={false} default="600.0">
  Request timeout in seconds
</ParamField>

<ParamField query="api_key" type="Optional" required={false}>
  Optional API key override
</ParamField>

<ParamField query="api_base" type="Optional" required={false}>
  Optional API base URL override
</ParamField>

<ParamField query="metadata" type="Optional" required={false}>
  Optional metadata for tracing \*\*kwargs: Additional arguments passed to litellm.embedding()
</ParamField>

### Returns

<ResponseField name="Returns" type="EmbeddingResult">
  EmbeddingResult with embeddings list, model, usage, and metadata
</ResponseField>

## Usage

```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
>>> from praisonaiagents import embedding
    >>> result = embedding("Hello, world!")
    >>> print(len(result.embeddings[0]))
    1536

    >>> result = embedding(["Hello", "World"], model="text-embedding-3-large")
    >>> print(len(result.embeddings))
    2
```

## Uses

* `litellm.embedding`
* `EmbeddingResult`

## Used By

* [`EmbeddingAgent.embed`](../functions/EmbeddingAgent-embed)
* [`EmbeddingAgent.embed_batch`](../functions/EmbeddingAgent-embed_batch)
* [`embedding`](../functions/embedding)
* [`Memory.search_short_term`](../functions/Memory-search_short_term)

## Source

<Card title="View on GitHub" icon="github" href="https://github.com/MervinPraison/PraisonAI/blob/main/src/praisonai-agents/praisonaiagents/embedding/embed.py#L13">
  `praisonaiagents/embedding/embed.py` at line 13
</Card>