Skip to main content
@ai-billing/xai
@ai-billing/xai / createXaiV3Middleware

​
Function: createXaiV3Middleware()

createXaiV3Middleware<TTags>(options): LanguageModelV3Middleware
Defined in: xai/src/ai-sdk/language-model-middleware/v3/language-model-v3-xai-billing-middleware.ts:98 Creates a V3 billing middleware for the xAI provider (@ai-sdk/xai). Deducts cache-read tokens from prompt tokens before billing — xAI charges only non-cached input at the prompt rate, and cached tokens separately at the cache-read rate.

​
Type Parameters

​
TTags

TTags extends JSONObject The shape of the tags object, extending DefaultTags.

​
Parameters

​
options

XaiV3MiddlewareOptions<TTags> Billing options; see XaiV3MiddlewareOptions. A priceResolver is required.

​
Returns

LanguageModelV3Middleware A V3 billing middleware instance for xAI.

​
Example

import { createXai } from '@ai-sdk/xai';
import { wrapLanguageModel } from 'ai';
import { createXAIMiddleware } from '@ai-billing/xai';
import {
  consoleDestination,
  createObjectPriceResolver,
  type ModelPricing,
} from '@ai-billing/core';

const xai = createXai({ apiKey: process.env.XAI_API_KEY });

const customPricingMap: Record<string, ModelPricing> = {
  'grok-3': {
    promptTokens: 3.0 / 1_000_000,
    completionTokens: 15.0 / 1_000_000,
    inputCacheReadTokens: 0.75 / 1_000_000,
  },
  'grok-3-mini': {
    promptTokens: 0.3 / 1_000_000,
    completionTokens: 0.5 / 1_000_000,
    internalReasoningTokens: 0.5 / 1_000_000,
    inputCacheReadTokens: 0.075 / 1_000_000,
  },
};

const priceResolver = createObjectPriceResolver(customPricingMap);

const billingMiddleware = createXAIMiddleware({
  destinations: [consoleDestination()],
  priceResolver,
});

const wrappedModel = wrapLanguageModel({
  model: xai('grok-3'),
  middleware: billingMiddleware,
});