voyage-ai-provider weekly downloads statistics

AI SDK - Voyage AI Provider

<div align="center"> <a href="https://www.npmjs.com/package/voyage-ai-provider"><img src="https://img.shields.io/npm/v/voyage-ai-provider"/><a> <a href="https://www.npmjs.com/package/voyage-ai-provider"><img src="https://img.shields.io/npm/dm/voyage-ai-provider"/><a> <a href="https://github.com/patelvivekdev/voyageai-ai-provider/actions/workflows/CI.yml"><img src="https://github.com/patelvivekdev/voyageai-ai-provider/actions/workflows/CI.yml/badge.svg"/><a> <a href="https://deepwiki.com/patelvivekdev/voyageai-ai-provider"><img src="https://deepwiki.com/badge.svg" alt="Ask DeepWiki"></a> </div> <br>

Introduction

The Voyage AI Provider is a provider for the AI SDK. It provides a simple interface to the Voyage AI API.

Version 5 targets AI SDK 7 and requires Node.js 22 or newer. It is distributed as an ESM-only package.

Installation

npm install voyage-ai-provider

# or

yarn add voyage-ai-provider

# or

pnpm add voyage-ai-provider

# or

bun add voyage-ai-provider

Configuration

The Voyage AI Provider requires an API key to be configured. You can obtain an API key by signing up at Voyage AI.

add the following to your .env file:

VOYAGE_API_KEY=your-api-key

Usage

Text Embedding

import { voyage } from 'voyage-ai-provider';
import { embedMany } from 'ai';

const embeddingModel = voyage.embeddingModel('voyage-3-lite');

export const generateEmbeddings = async (
  value: string,
): Promise<Array<{ embedding: number[]; content: string }>> => {
  // Generate chunks from the input value
  const chunks = value.split('\n');

  // Optional: You can also split the input value by comma
  // const chunks = value.split('.');

  // Or you can use LLM to generate chunks(summarize) from the input value

  const { embeddings } = await embedMany({
    model: embeddingModel,
    values: chunks,
  });
  return embeddings.map((e, i) => ({ content: chunks[i], embedding: e }));
};

How to pass additional settings to the model

The settings object should contain the settings you want to add to the model. You can find the available settings for the model in the Voyage API documentation: https://docs.voyageai.com/reference/embeddings-api

import { embedMany } from 'ai';
import { createVoyage } from 'voyage-ai-provider';

const voyage = createVoyage({
  apiKey: process.env.VOYAGE_API_KEY,
});

const embeddingModel = voyage.embeddingModel('voyage-3-lite');

const { embeddings } = await embedMany({
  model: embeddingModel,
  values: ['sunny day at the beach', 'rainy day in the city'],
  providerOptions: {
    voyage: {
      inputType: 'document',
      outputDimension: 1024,
      outputDtype: 'float',
    },
  },
});

Image & Multi-modal Embedding

Multimodal and image embeddings both use the voyage-multimodal-3 model and the same /multimodalembeddings endpoint. Following the AI SDK convention (and the official providers such as Google), the embed/embedMany values array holds the text for each embedding, and any non-text content (images) is passed via providerOptions.voyage.content.

content is an array aligned to values by index: content[i] are the extra parts merged with the text in values[i]. Its length must equal values.length. Use null for entries that are text-only. For an image-only embedding, pass an empty string ('') for that value.

Each content part is one of:

• { type: 'text', text: string }
• { type: 'image_url', image_url: string }
• { type: 'image_base64', image_base64: string }

Example 1: A single image per embedding (image-only)

import {
  voyage,
  type VoyageMultimodalEmbeddingOptions,
} from 'voyage-ai-provider';
import { embedMany } from 'ai';

const imageModel = voyage.imageEmbeddingModel('voyage-multimodal-3');

const { embeddings } = await embedMany({
  model: imageModel,
  values: ['', ''], // one empty string per image-only embedding
  providerOptions: {
    voyage: {
      content: [
        [
          {
            type: 'image_url',
            image_url:
              'https://raw.githubusercontent.com/voyage-ai/voyage-multimodal-3/refs/heads/main/images/banana_200_x_200.jpg',
          },
        ],
        [
          {
            type: 'image_base64',
            image_base64: 'data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAA...',
          },
        ],
      ],
    } satisfies VoyageMultimodalEmbeddingOptions,
  },
});

Example 2: Multiple images in a single embedding

import {
  voyage,
  type VoyageMultimodalEmbeddingOptions,
} from 'voyage-ai-provider';
import { embedMany } from 'ai';

const imageModel = voyage.imageEmbeddingModel('voyage-multimodal-3');

const { embeddings } = await embedMany({
  model: imageModel,
  values: [''],
  providerOptions: {
    voyage: {
      content: [
        [
          {
            type: 'image_url',
            image_url:
              'https://raw.githubusercontent.com/voyage-ai/voyage-multimodal-3/refs/heads/main/images/banana_200_x_200.jpg',
          },
          {
            type: 'image_base64',
            image_base64: 'data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAA...',
          },
        ],
      ],
    } satisfies VoyageMultimodalEmbeddingOptions,
  },
});

Example 3: Text and images combined per embedding

import {
  voyage,
  type VoyageMultimodalEmbeddingOptions,
} from 'voyage-ai-provider';
import { embedMany } from 'ai';

const multimodalModel = voyage.multimodalEmbeddingModel('voyage-multimodal-3');…