{
  "id": 11836236,
  "title": "Grok 4.7 pricing: same $2/$6 per token, double the cost per task",
  "url": "https://urgent.news/2026/10/04/grok-4-7-pricing-same-2-6-per-token-double-the-cost-per-task",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-04T03:46:44.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/axrisi/grok-47-pricing-same-26-per-token-double-the-cost-per-task-43d8"
  },
  "original_language": "en",
  "account": "Grok 4.7 pricing remains unchanged at $2 per million input tokens and $6 per million output tokens, though the cost per task has doubled compared to its predecessor. Artificial Analysis found that Grok 4.7 generates approximately 2.5 times as many tokens as Grok 4.6, resulting in a cost increase from $1.86 to $3.74 per task. Despite the unchanged price per token, the model's output speed has decreased to 57 tokens per second, making it notably slow and verbose. While it performs well on long, multi-step tasks and coding, it has also been reported as lazy and dishonest by some developers.",
  "summary": "Grok 4.7 costs exactly what Grok 4.6 did per token: $2 per million input tokens and $6 per million output. Per task, it costs about double. Artificial Analysis measured xAI's new model writing roughly 2.5 times as many tokens as its predecessor, so its cost per task went from $1.86 to $3.74. If you pick LLMs by the price table, this is the week that table stopped telling you what you will pay.…",
  "key_points": [
    "Grok 4.7 pricing unchanged at $2/$6 per million tokens",
    "Task cost doubled to $3.74 from $1.86 per task",
    "Output speed reduced to 57 tokens per second"
  ],
  "editors_take": "The unchanged per-token pricing for Grok 4.7 belies a doubling of cost per task due to increased token generation, affecting developers relying on the model's performance and speed.",
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}