> ## Documentation Index
> Fetch the complete documentation index at: https://narev.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Getting started

> First steps with GMI Cloud middleware for tracking AI costs.

## Installation

<CodeGroup>
  ```bash npm theme={null}
  npm install @ai-billing/gmicloud @ai-billing/core @ai-sdk/gmicloud ai
  ```

  ```bash pnpm theme={null}
  pnpm add @ai-billing/gmicloud @ai-billing/core @ai-sdk/gmicloud ai
  ```

  ```bash yarn theme={null}
  yarn add @ai-billing/gmicloud @ai-billing/core @ai-sdk/gmicloud ai
  ```
</CodeGroup>

## Overview

The `@ai-billing/gmicloud` package provides middleware for tracking token usage and calculating costs when using GMI Cloud models with the Vercel AI SDK.

It captures GMI Cloud-specific metrics, such as `completion_tokens_details.reasoning_tokens` and `prompt_tokens_details.cached_tokens`, ensuring that reasoning and prompt-caching costs are accurately reflected.

## Usage

To use the middleware, wrap your GMI Cloud model using `wrapLanguageModel` from the `ai` package and pass the `createGmicloudMiddleware`.

<Steps>
  <Step title="Initialize the GMI Cloud provider">
    First, set up the GMI Cloud provider using your API key.

    ```typescript theme={null}
    import { createGmicloud } from '@ai-sdk/gmicloud';

    const gmicloud = createGmicloud({
      apiKey: process.env.GMI_CLOUD_APIKEY,
    });
    ```
  </Step>

  <Step title="Define model pricing">
    Set up a price resolver to define the costs for the models you'll be using. For GMI Cloud, you can specify costs for standard prompt/completion tokens and reasoning tokens.

    ```typescript theme={null}
    import { createObjectPriceResolver } from '@ai-billing/core';

    const priceResolver = createObjectPriceResolver({
      'deepseek-ai/DeepSeek-V4-Flash-0731': {
        promptTokens: 0.0000001,
        completionTokens: 0.0000003,
      },
    });
    ```
  </Step>

  <Step title="Create the billing middleware">
    Initialize the GMI Cloud billing middleware. You need to provide a destination (such as `consoleDestination`) where billing events will be sent, along with your `priceResolver`.

    ```typescript theme={null}
    import { createGmicloudMiddleware } from '@ai-billing/gmicloud';
    import { consoleDestination } from '@ai-billing/core';

    const billingMiddleware = createGmicloudMiddleware({
      destinations: [consoleDestination()],
      priceResolver: priceResolver,
    });
    ```
  </Step>

  <Step title="Wrap the model">
    Use `wrapLanguageModel` from the `ai` package to apply the billing middleware to your GMI Cloud model.

    ```typescript theme={null}
    import { wrapLanguageModel } from 'ai';

    const wrappedModel = wrapLanguageModel({
      model: gmicloud('deepseek-ai/DeepSeek-V4-Flash-0731'),
      middleware: billingMiddleware,
    });
    ```
  </Step>

  <Step title="Use the wrapped model">
    Finally, use the wrapped model with AI SDK functions like `generateText` or `streamText`. The billing middleware will automatically track tokens, handle reasoning and caching metrics, and calculate costs.

    ```typescript theme={null}
    import { generateText } from 'ai';

    const result = await generateText({
      model: wrappedModel,
      prompt: 'What is the capital of Sweden?',
    });
    ```
  </Step>
</Steps>