Embeddings · flua

Text vectors for search, RAG and recommendations — OpenAI and Google models in the OpenAI format.

POST/v1/embeddings

An embedding is a vector of numbers that captures what a text means: texts with similar meaning get nearby vectors. Semantic search, RAG, recommendations and clustering are built on them.

Request#

import os
from openai import OpenAI

client = OpenAI(base_url="https://api.flua.ink/v1", api_key=os.environ["FLUA_API_KEY"])

result = client.embeddings.create(model="text-embedding-3-small", input="Swifts spend almost their whole life in flight")
print(len(result.data[0].embedding))

Parameter	Type	Description
`model`*	string	The embedding model.
`input`*	string \| string[]	A text or an array of texts — up to 2048 per request.
`dimensions`	integer	Shorten the vector (text-embedding-3-*): less storage, slightly lower accuracy.
`encoding_format`	string	float (default) or base64.

Response#

200 OK

{
  "object": "list",
  "model": "text-embedding-3-small",
  "data": [
    { "object": "embedding", "index": 0, "embedding": [0.0123, -0.0456, …] }
  ],
  "usage": { "prompt_tokens": 9, "total_tokens": 9 }
}

Models#

Text Embedding 3 Large — text-embedding-3-large
Text Embedding 3 Small — text-embedding-3-small
Text Embedding Ada 002 — text-embedding-ada-002
Gemini Embedding 2 Preview — gemini-embedding-2-preview

Embeddings are billed for input tokens only. Vectors of different models are not compatible: switch models and you must re-embed everything — which is why fallback models never apply to embeddings.