{
  "schemaVersion": 1,
  "id": "gemini-4-argon",
  "provider": "google",
  "lastVerifiedAt": "2026-10-16T10:37:03.951053Z",
  "sourceUrls": [
    "https://ai.google.dev/gemini-api/docs/pricing",
    "https://ai.google.dev/gemini-api/docs/models",
    "https://ai.google.dev/gemini-api/docs/models/gemini-4-argon",
    "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/",
    "https://storage.googleapis.com/deepmind-media/gemini/gemini_4_safety_report.pdf"
  ],
  "availability": "available",
  "publicApiModelId": "gemini-4-argon",
  "accessNote": "Google now lists the stable gemini-4-argon endpoint for paid Gemini API access. Not available on the Free tier or Google AI Pro plan; project quotas and regional eligibility apply.",
  "pricing": {
    "inputPerM": 2,
    "cachedInputPerM": 0.1,
    "outputPerM": 10,
    "outputIncludesThinkingTokens": true,
    "cacheMode": "implicit-only",
    "introductoryEndsOn": "2026-11-30",
    "afterIntroductory": {
      "effectiveDate": "2026-12-01",
      "inputPerM": 4,
      "cachedInputPerM": 0.2,
      "outputPerM": 20,
      "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
      "verifiedAt": "2026-10-16T10:35:21.725778Z"
    },
    "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
    "verifiedAt": "2026-10-16T10:37:03.951053Z"
  },
  "scheduledPricing": {
    "effectiveDate": "2026-12-01",
    "effectiveTimeZone": null,
    "baselineLabel": "Current introductory rate",
    "tiers": [
      {
        "name": "Standard rate from December 1, 2026",
        "pricing": {
          "inputPerM": 4,
          "cachedInputPerM": 0.2,
          "outputPerM": 20
        }
      }
    ],
    "sourceUrl": "https://ai.google.dev/gemini-api/docs/pricing",
    "verifiedAt": "2026-10-16T10:37:03.951053Z",
    "note": "Google publishes a date, not an exact transition time or timezone. Introductory rates run through November 30, 2026. Implicit caching only."
  },
  "context": 1048576,
  "maxOutput": 1048576,
  "maxOutputLongDecode": 1048576,
  "caching": {
    "implicit": true,
    "explicit": false,
    "sourceUrl": "https://ai.google.dev/gemini-api/docs/models/gemini-4-argon",
    "verifiedAt": "2026-10-16T10:37:03.951053Z"
  },
  "researchClaim": {
    "claim": "Google researchers use models to help develop successor models",
    "primaryEvidence": "https://storage.googleapis.com/deepmind-media/gemini/gemini_4_safety_report.pdf",
    "boundary": "Google reports human-reviewed code assistance and continued human research direction; not proof of fully autonomous recursive self-improvement or lower customer token prices.",
    "brinQuoteSource": "https://www.businessinsider.com/sergey-brin-google-gemini-ai-argon-recursive-self-improvement-rsi-2026-10",
    "brinQuoteDirectlyVerified": false,
    "newModelOrPriceCut": false
  },
  "labs": {
    "status": "blocked-pending-authorized-evaluation",
    "reason": "Public paid endpoint is documented, but no authorized funded Argon evaluation with this exact route, fixed reasoning level and full response/token/retry/latency/billing ledger has been completed by AI Pricing Guru. No paid calls initiated for this release.",
    "unblockRequirements": [
      "Authorized funded exact-model access and budget",
      "Pinned endpoint/settings and frozen task set",
      "Raw outputs, accepted task scores, tokens including thinking, retries, latency and billed cost"
    ],
    "rsiExperiment": "A multi-generation research-automation claim additionally needs controlled baselines, human-intervention logs and end-to-end compute costs; the 49-task leaderboard cannot establish recursive self-improvement."
  },
  "articlePath": "/news/google-gemini-self-improvement-brin-pricing-october-2026/",
  "providerPath": "/google-ai-pricing/",
  "maintenance": {
    "owner": "aipricing-data",
    "cadenceDays": 1,
    "watch": "Daily pricing pipeline plus pricing-asset official-source and freshness watch; review access, price dates, schedule promotion and claim boundaries; never advance verification without checking sources."
  },
  "defaultMaxOutput": 65536
}