Implementation:Cohere ai Cohere python EmbedFloatsResponse Model

Knowledge Sources	Cohere Python SDK
Domains	SDK, Embeddings
Last Updated	2026-02-15 14:00 GMT

Overview

EmbedFloatsResponse is a Pydantic model representing the Cohere Embed API response containing embeddings exclusively as arrays of floats.

Description

The EmbedFloatsResponse class encapsulates the response from the Cohere Embed API when embeddings are returned in the default float-only format. Unlike EmbedByTypeResponse which supports multiple numeric formats, this response contains a single embeddings field as a nested list of floats (List[List[float]]).

The response includes:

• id: A unique identifier for the embed request
• embeddings: The embedding vectors as a list of float arrays, where each inner list corresponds to one input text or image
• texts: The original text entries that were embedded
• images: The image entries that were embedded (if any)
• meta: Optional API metadata including token counts and warnings

The length of the embeddings array matches the number of input texts provided. The class extends UncheckedBaseModel and is auto-generated by the Fern API definition toolchain.

Usage

Use EmbedFloatsResponse when working with the default float embedding response from the Cohere Embed API. This is the standard response type when no specific embedding_types parameter is provided or when only float embeddings are needed.

Code Reference

Source Location

• Repository: Cohere Python SDK
• File: src/cohere/types/embed_floats_response.py

Signature

class EmbedFloatsResponse(UncheckedBaseModel):
    id: str
    embeddings: typing.List[typing.List[float]]
    texts: typing.List[str]
    images: typing.Optional[typing.List[Image]] = None
    meta: typing.Optional[ApiMeta] = None

Import

from cohere.types import EmbedFloatsResponse

I/O Contract

Fields

Field	Type	Required	Default	Description
id	str	Yes	--	Unique identifier for this embed request
embeddings	List[List[float]]	Yes	--	An array of embeddings where each embedding is an array of floats; length matches the input array
texts	List[str]	Yes	--	The text entries for which embeddings were returned
images	Optional[List[Image]]	No	None	The image entries for which embeddings were returned
meta	Optional[ApiMeta]	No	None	API metadata including token counts and warnings

Usage Examples

Basic Float Embedding Response

import cohere

co = cohere.Client()

response = co.embed(
    texts=["Hello world", "Embeddings are useful"],
    model="embed-english-v3.0",
    input_type="search_document",
)

# Access the embeddings
print(f"Request ID: {response.id}")
print(f"Number of embeddings: {len(response.embeddings)}")
print(f"Embedding dimension: {len(response.embeddings[0])}")

# Iterate over text-embedding pairs
for text, embedding in zip(response.texts, response.embeddings):
    print(f"Text: {text[:30]}... -> Embedding dim: {len(embedding)}")

Computing Cosine Similarity

import cohere
import numpy as np

co = cohere.Client()

response = co.embed(
    texts=["machine learning", "artificial intelligence", "cooking recipes"],
    model="embed-english-v3.0",
    input_type="search_document",
)

# Convert to numpy arrays for similarity computation
embeddings = np.array(response.embeddings)

# Cosine similarity between first two texts
similarity = np.dot(embeddings[0], embeddings[1]) / (
    np.linalg.norm(embeddings[0]) * np.linalg.norm(embeddings[1])
)
print(f"Similarity between '{response.texts[0]}' and '{response.texts[1]}': {similarity:.4f}")

Related Pages

• Environment:Cohere_ai_Cohere_python_Python_SDK_Runtime