wrapLanguageModel()

The wrapLanguageModel function provides a way to enhance the behavior of language models by wrapping them with middleware. See Language Model Middleware for more information on middleware.

import { wrapLanguageModel, gateway } from 'ai';

const wrappedLanguageModel = wrapLanguageModel({
  model: gateway('openai/gpt-4.1'),
  middleware: yourLanguageModelMiddleware,
});

Import

<Snippet text={import { wrapLanguageModel } from "ai"} prompt={false} />

API Signature

Parameters

<PropertiesTable content={[ { name: 'model', type: 'LanguageModelV4', description: 'The original LanguageModelV4 instance to be wrapped.', }, { name: 'middleware', type: 'LanguageModelV4Middleware | LanguageModelV4Middleware[]', description: 'The middleware to be applied to the language model. When multiple middlewares are provided, the first middleware will transform the input first, and the last middleware will be wrapped directly around the model.', }, { name: 'modelId', type: 'string', description: "Optional custom model ID to override the original model's ID.", }, { name: 'providerId', type: 'string', description: "Optional custom provider ID to override the original model's provider.", }, ]} />

Returns

A new LanguageModelV4 instance with middleware applied.