{
  "metadata": {
    "name": "Effective AI Consumer Price-per-Task: An Exploratory Computation (Claude Pro, ChatGPT Plus, ChatGPT Pro), Jan vs Jun 2026",
    "description": "An exploratory computation of effective AI consumer price per task across three major subscription plans (Claude Pro, ChatGPT Plus, ChatGPT Pro), comparing Jan 2026 vs Jun 2026. This is a small sample (n=3 plans, one of which has no January value), not a comprehensive index. Effective $/task is computed from a fixed task basket priced at each plan's list price, then adjusted for tokenizer inflation and free-tier changes; it is a computed comparison metric, NOT a measured bill. The comparison illustrates list-price-flat 'shrinkflation' driven by the Opus 4.7 tokenizer (carried into Opus 4.8).",
    "tools": 3,
    "timeRange": "2026-01 to 2026-06",
    "datePublished": "2026-06-10",
    "dateModified": "2026-06-10",
    "nextReview": "2026-09",
    "license": "CC-BY-4.0",
    "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
    "canonicalUrl": "https://nesyona.com/articles/real-ai-price-shrinkflation-index-2026",
    "creator": {
      "name": "Vincent Wesley Couey",
      "id": "https://nesyona.com/author/vincent-couey#person"
    },
    "publisher": {
      "name": "Nesyona",
      "url": "https://nesyona.com"
    },
    "methodology": "Fixed task basket of 300 tasks/month (avg 850 input + 600 output tokens each), priced at each tool's current list price, multiplied by its tokenizer-inflation factor, then adjusted for free-tier headroom (fraction of the basket the free tier covers). The result is effective $/task: what a moderate-use consumer actually spends per meaningful output task at the current tokenization rate. Tokenizer-inflation factors sourced from OpenRouter, FindSkill, and TokenMix analyses of the Claude Opus 4.7 tokenizer. Every figure is date-stamped and source-attributed.",
    "referenceBasket": {
      "tasksPerMonth": 300,
      "avgInputTokens": 850,
      "avgOutputTokens": 600
    },
    "fields": [
      "tool",
      "list_price_usd_per_month",
      "tokenizer_inflation_factor",
      "effective_dollars_per_task_jan_2026",
      "effective_dollars_per_task_jun_2026",
      "change",
      "sources",
      "verified_date"
    ],
    "dataProvenance": "SOURCED: each tool's current list price (read from the vendor's public pricing page). DERIVED/COMPUTED: the effective-price figure = a fixed 300-task/month basket (avg 850 input + 600 output tokens/task) priced at list, multiplied by each tool's tokenizer-inflation factor, adjusted for free-tier headroom. This is a computed comparison metric across three plans (a small, exploratory sample), NOT a measured bill and NOT a comprehensive index."
  },
  "data": [
    {
      "tool": "Claude Pro",
      "list_price_usd_per_month": 20,
      "tokenizer_inflation_factor": "1.0-1.35x",
      "effective_dollars_per_task_jan_2026": 0.067,
      "effective_dollars_per_task_jun_2026": 0.085,
      "change": "+27% max (12-27% cost impact; +32-47% more tokens)",
      "sources": [
        "OpenRouter",
        "FindSkill",
        "TokenMix"
      ],
      "verified_date": "2026-06-10"
    },
    {
      "tool": "ChatGPT Plus",
      "list_price_usd_per_month": 20,
      "tokenizer_inflation_factor": "1.0x",
      "effective_dollars_per_task_jan_2026": 0.061,
      "effective_dollars_per_task_jun_2026": 0.066,
      "change": "+8% (tier-value; free tier ad-degraded Feb 2026)",
      "sources": [
        "FindSkill",
        "Fritz.ai",
        "CloudZero"
      ],
      "verified_date": "2026-06-10"
    },
    {
      "tool": "ChatGPT Pro",
      "list_price_usd_per_month": 100,
      "tokenizer_inflation_factor": "1.0x",
      "effective_dollars_per_task_jan_2026": null,
      "effective_dollars_per_task_jun_2026": 0.04,
      "change": "new tier Apr 2026",
      "sources": [
        "BuildFastWithAI"
      ],
      "verified_date": "2026-06-10"
    }
  ],
  "keyFindings": [
    "Claude Pro's effective cost per task rose approximately 20-27% in 2026 with no list-price change, driven by Opus 4.7 tokenizer inflation (April 2026): the same prompt now generates 12-47% more tokens.",
    "ChatGPT Plus list price held at $20/mo but effective value eroded ~8% via Feb 2026 free-tier ad-degradation and GPT-5.5 being gated to the $100+ Pro tiers.",
    "The list price is a fiction; the effective price is the number a user feels in their usage cap."
  ]
}