Mixedbread Provider

patelvivekdev/mixedbread-ai-provider is a community provider that uses Mixedbread to provide Embedding support for the AI SDK.

Setup

The Mixedbread provider is available in the mixedbread-ai-provider module. You can install it with

pnpm add mixedbread-ai-provider

Provider Instance

You can import the default provider instance mixedbread from mixedbread-ai-provider:

import { mixedbread } from 'mixedbread-ai-provider';

If you need a customized setup, you can import createMixedbread from mixedbread-ai-provider and create a provider instance with your settings:

import { createMixedbread } from 'mixedbread-ai-provider';

const mixedbread = createMixedbread({
  // custom settings
});

You can use the following optional settings to customize the Mixedbread provider instance:

• baseURL string

  The base URL of the Mixedbread API. The default prefix is https://api.mixedbread.com/v1.

• apiKey string

  API key that is being sent using the Authorization header. It defaults to the MIXEDBREAD_API_KEY environment variable.

• headers Record<string,string>

  Custom headers to include in the requests.

• fetch (input: RequestInfo, init?: RequestInit) => Promise<Response>

  Custom fetch implementation. Defaults to the global fetch function. You can use it as a middleware to intercept requests, or to provide a custom fetch implementation for e.g. testing.

Text Embedding Models

You can create models that call the Mixedbread embeddings API using the .textEmbeddingModel() factory method.

import { mixedbread } from 'mixedbread-ai-provider';

const embeddingModel = mixedbread.textEmbeddingModel(
  'mixedbread-ai/mxbai-embed-large-v1',
);

You can use Mixedbread embedding models to generate embeddings with the embed function:

import { mixedbread } from 'mixedbread-ai-provider';
import { embed } from 'ai';

const { embedding } = await embed({
  model: mixedbread.textEmbeddingModel('mixedbread-ai/mxbai-embed-large-v1'),
  value: 'sunny day at the beach',
});

Mixedbread embedding models support additional provider options that can be passed via providerOptions.mixedbread:

import { mixedbread } from 'mixedbread-ai-provider';
import { embed } from 'ai';

const { embedding } = await embed({
  model: mixedbread.textEmbeddingModel('mixedbread-ai/mxbai-embed-large-v1'),
  value: 'sunny day at the beach',
  providerOptions: {
    mixedbread: {
      prompt: 'Generate embeddings for text',
      normalized: true,
      dimensions: 512,
      encodingFormat: 'float16',
    },
  },
});

The following provider options are available:

• prompt string

  An optional prompt to provide context to the model. Refer to the model's documentation for more information. A string between 1 and 256 characters.

• normalized boolean

  Option to normalize the embeddings.

• dimensions number

  The desired number of dimensions in the output vectors. Defaults to the model's maximum. A number between 1 and the model's maximum output dimensions. Only applicable for Matryoshka-based models.

• encodingFormat 'float' | 'float16' | 'binary' | 'ubinary' | 'int8' | 'uint8' | 'base64'

Model Capabilities