Implementation:Helicone Helicone XAI Endpoint Pricing

Knowledge Sources	Helicone
Domains	Cost Calculation, Model Pricing
Last Updated	2026-02-14 06:32 GMT

Overview

Defines endpoint pricing configurations for xAI Grok models across xAI direct, Helicone gateway, and OpenRouter providers, including tiered pricing based on context window size.

Description

This file exports an endpoints object that maps xAI Grok model endpoint identifiers (in the format model-name:provider) to their full configuration including per-token pricing, context length limits, max completion tokens, cache multipliers, web search fees, and supported API parameters. The file covers the complete Grok model family: Grok 3, Grok 3 Mini, Grok 4, Grok 4 Fast (reasoning and non-reasoning variants), Grok 4.1 Fast, and Grok Code Fast 1. Each model is defined across up to three providers (xAI direct, Helicone gateway with pa/ prefix, and OpenRouter with x-ai/ prefix).

A key feature is tiered pricing: several models (Grok 4, Grok 4 Fast Reasoning, Grok 4.1 Fast Reasoning) have two pricing tiers based on context window usage, with a threshold at 128,000 tokens. Input above this threshold is billed at higher rates (typically 2x).

Usage

This file is consumed by Helicone's cost calculation engine to compute the cost of xAI Grok model requests. When a request is logged through the proxy, the cost system looks up the model's endpoint configuration and applies the appropriate per-token pricing tier based on the number of input tokens used.

Code Reference

Source Location

• Repository: Helicone
• File: packages/cost/models/authors/xai/endpoints.ts

Signature

import { ModelProviderName } from "../../providers";
import type { ModelProviderConfig } from "../../types";
import { GrokModelName } from "./models";

export const endpoints = {
  "grok-code-fast-1:xai": { ... },
  "grok-4:xai": { ... },
  "grok-4-fast-reasoning:xai": { ... },
  // ... more endpoints
} satisfies Partial<
  Record<`${GrokModelName}:${ModelProviderName}`, ModelProviderConfig>
>;

Import

import { endpoints } from "@helicone-package/cost/models/authors/xai/endpoints";

Endpoint Configuration Schema

Each endpoint entry follows the ModelProviderConfig type:

{
  providerModelId: string,          // Provider-specific model identifier
  provider: string,                 // Provider name (xai, helicone, openrouter)
  author: "xai",                    // Model author
  providerModelIdAliases?: string[], // Alternative model ID strings
  pricing: [
    {
      threshold: number,            // Token count threshold for tier (0 = base tier)
      input: number,                // Cost per input token
      output: number,               // Cost per output token
      request?: number,             // Per-request cost
      web_search?: number,          // Cost per web search source
      cacheMultipliers?: {
        cachedInput: number,        // Multiplier for cached input tokens
      },
    },
  ],
  contextLength: number,            // Maximum context window size
  maxCompletionTokens: number,      // Maximum output tokens
  supportedParameters: string[],    // Supported API parameters
  ptbEnabled: boolean,              // Pass-through billing enabled
  endpointConfigs: { "*": {} },     // Endpoint configuration (wildcard)
}

Model Endpoints

xAI Direct Provider

Endpoint Key	Model ID	Input (per 1M)	Output (per 1M)	Context	Tiered
grok-code-fast-1:xai	grok-code-fast-1	$0.20	$1.50	256K	No
grok-4:xai	grok-4	$3.00 / $6.00	$15.00 / $30.00	256K	Yes (128K)
grok-4-fast-reasoning:xai	grok-4-fast	$0.20 / $0.40	$0.50 / $1.00	2M	Yes (128K)
grok-4-fast-non-reasoning:xai	grok-4-fast-non-reasoning	$0.20	$0.50	2M	No
grok-4-1-fast-non-reasoning:xai	grok-4-1-fast-non-reasoning	$0.20	$0.50	2M	No
grok-4-1-fast-reasoning:xai	grok-4-1-fast-reasoning	$0.20 / $0.40	$0.50 / $1.00	2M	Yes (128K)
grok-3:xai	grok-3	$3.00	$15.00	131K	No
grok-3-mini:xai	grok-3-mini	$0.30	$0.50	131K	No

Helicone Gateway Provider

Mirrors xAI direct pricing but with pa/ prefixed model IDs for passthrough billing:

Endpoint Key	Model ID	Pricing
grok-4:helicone	pa/grk-4	Same as grok-4:xai
grok-4-fast-reasoning:helicone	pa/grok-4-fast-reasoning	Same as grok-4-fast-reasoning:xai
grok-4-fast-non-reasoning:helicone	pa/grok-4-fast-non-reasoning	Same as grok-4-fast-non-reasoning:xai
grok-4-1-fast-non-reasoning:helicone	pa/grok-4-1-fast-non-reasoning	Same pricing
grok-4-1-fast-reasoning:helicone	pa/grok-4-1-fast-reasoning	Same pricing
grok-code-fast-1:helicone	pa/grok-code-fast-1	Same pricing
grok-3:helicone	pa/grk-3	Same as grok-3:xai
grok-3-mini:helicone	pa/grok-3-mini	Same as grok-3-mini:xai

OpenRouter Provider

Uses x-ai/ prefixed model IDs with slightly higher pricing (approximately 5.5% markup):

Endpoint Key	Model ID	Input (per 1M)	Output (per 1M)
grok-3:openrouter	x-ai/grok-3	$5.28	$26.38
grok-3-mini:openrouter	x-ai/grok-3-mini	$0.63	$4.22
grok-4:openrouter	x-ai/grok-4	$6.33	$31.65
grok-code-fast-1:openrouter	x-ai/grok-code-fast-1	$0.21	$1.58

I/O Contract

Inputs

Name	Type	Required	Description
GrokModelName	string union	Yes	One of the defined Grok model names (imported from ./models)
ModelProviderName	string union	Yes	Provider identifier (xai, helicone, openrouter)

Outputs

Name	Type	Description
endpoints	Record	Map of endpoint keys to ModelProviderConfig objects

Usage Examples

import { endpoints } from "./endpoints";

// Look up pricing for Grok 4 on xAI
const grok4Config = endpoints["grok-4:xai"];
// grok4Config.pricing[0].input = 0.000003 ($3.00 per 1M tokens, up to 128K)
// grok4Config.pricing[1].input = 0.000006 ($6.00 per 1M tokens, over 128K)

// Check web search cost
const webSearchCost = grok4Config.pricing[0].web_search;
// 0.025 ($25.00 per 1K sources)

// Get cache discount
const cacheMultiplier = grok4Config.pricing[0].cacheMultipliers?.cachedInput;
// 0.25 (75% discount on cached input)

Related Pages

• Principle:Helicone_Helicone_Endpoint_Pricing_Configuration