Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/langchain-ai/langchain/llms.txt

Use this file to discover all available pages before exploring further.

The NomicEmbeddings class provides integration with Nomic’s embedding models, supporting both text and image embeddings.

​
Installation

pip install langchain-nomic

​
Setup

Set your Nomic API key: 
export NOMIC_API_KEY="your-api-key"

​
Usage

​
Basic usage

from langchain_nomic import NomicEmbeddings

embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5"
)

​
Embed single text

text = "The meaning of life is 42"
vector = embed.embed_query(text)
print(len(vector))

​
Embed multiple texts

texts = ["hello world", "goodbye world"]
vectors = embed.embed_documents(texts)
print(len(vectors))

​
Configuration

​
Supported models

Text models:
  • nomic-embed-text-v1.5 - Latest text embedding model (768 dimensions)
  • nomic-embed-text-v1 - Previous generation text model
Vision models:
  • nomic-embed-vision-v1.5 - Latest vision embedding model

​
Inference modes

Nomic supports three inference modes:

​
Remote inference (default)

embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    inference_mode="remote"
)

​
Local inference

Run embeddings locally using Embed4All: 
embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    inference_mode="local",
    device="cpu"  # or "gpu"
)

​
Dynamic inference

Automatically choose between local and remote: 
embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    inference_mode="dynamic"
)

​
Matryoshka dimensions

Reduce embedding dimensions for faster search: 
embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    dimensionality=256  # Reduce from 768 to 256
)

​
Image embeddings

from langchain_nomic import NomicEmbeddings

embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    vision_model="nomic-embed-vision-v1.5"
)

# Embed images by URI
image_uris = [
    "https://example.com/image1.jpg",
    "file:///path/to/image2.png"
]
vectors = embed.embed_image(image_uris)

​
Device selection (local mode)

Do not use the device parameter on macOS.
For local inference on Linux/Windows: 
embed = NomicEmbeddings(
    model="nomic-embed-text-v1.5",
    inference_mode="local",
    device="nvidia"  # or "cpu", "gpu", "amd"
)

​
Parameters

​
model
string
required
Name of the Nomic text embedding model to use.
​
nomic_api_key
string
Nomic API key. Automatically inferred from NOMIC_API_KEY environment variable if not provided.
​
dimensionality
integer
Embedding dimension for Matryoshka-capable models. Defaults to full-size (768).
​
inference_mode
string
default:"remote"
How to generate embeddings. Options: remote, local, or dynamic.
​
device
string
Device for local embeddings. Options: cpu, gpu, nvidia, amd, or specific device name.
​
vision_model
string
Vision model to use for image embeddings.

Build docs developers (and LLMs) love

Get started for freeTalk to us