{
  "id": 5583096,
  "title": "Frontier AI faces pricing reckoning as token volume explodes 25-fold — mid-tier models deliver 90% of flagship capability at one-sixth the cost",
  "url": "https://urgent.news/2026/09/04/frontier-ai-faces-pricing-reckoning-as-token-volume-explodes-25-fold",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-04T15:21:56.000Z",
  "source": {
    "name": "Tom's Hardware",
    "slug": "tom-s-hardware",
    "url": "https://www.tomshardware.com/tech-industry/artificial-intelligence/frontier-ai-faces-pricing-reckoning-as-token-volume-explodes-25-fold-mid-tier-models-deliver-90-percent-of-flagship-capability-at-one-sixth-the-cost"
  },
  "original_language": "en",
  "account": "The frontier of artificial intelligence faces a pricing revolution as token volume has surged 25-fold in the past year, according to recent reports. Mid-tier models are delivering 90% of flagship capabilities at a fraction of the cost. While Anthropic's Claude Fable and Opus models have often been at the top of intelligence charts, they are also the most expensive to use. However, the competition is fierce among models that offer peak intelligence and minimal cost. Google's Gemini 3.8 Flash, for example, has taken the title from Google's Gemini 3.8 Flash and held it for only 3.5 hours. This shift in the AI landscape is causing companies to rethink their token usage, with overall paid token use increasing by 25 times in the past year and doubling in the past month alone. The focus on cost-effectiveness is leading to a migration towards the Pareto Frontier, where the best models are able to offer top-tier intelligence at a much lower cost.",
  "summary": "As frontier AI developers push for cost savings as much as intelligence enhancements, new models push the boundaries of the pareto frontier, with even small advantages crowning new kings.",
  "key_points": [
    "Token volume increased 25-fold in the past year",
    "Mid-tier models deliver 90% of flagship capabilities at one-sixth the cost",
    "Google's Gemini 3.8 Flash took top title for 3.5 hours"
  ],
  "editors_take": null,
  "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."
}