{
  "post": {
    "title": "Prompt caching, batching and model routing: what they actually save",
    "url": "https://bardak.ai/posts/prompt-caching-batching-routing",
    "updated": "2026-11-19T07:00:26.880Z"
  },
  "citation": "Prompt caching, batching and model routing: what they actually save, Ivan Bardak, bardak.ai, https://bardak.ai/posts/prompt-caching-batching-routing",
  "datasets": [
    {
      "key": "caching-rules-by-provider",
      "title": "Caching rules by provider",
      "type": "benchmarkTable",
      "source": "Anthropic, OpenAI and Google caching, batch and pricing documentation, read 16 November 2026",
      "caption": "Read fractions are of the standard input price. Older OpenAI models (GPT-5.5 and earlier) have no write charge. Gemini 4 models cache only implicitly, and Google calls that a discount with no guarantee.",
      "csv": "https://bardak.ai/data/prompt-caching-batching-routing/caching-rules-by-provider.csv",
      "header": [
        "Provider",
        "Reading from cache",
        "Writing to cache",
        "Smallest prompt",
        "Lifetime",
        "On by default",
        "With batch"
      ],
      "rows": [
        [
          "Anthropic",
          "A tenth of the input price; a twentieth on Opus 5.5 and Sonnet 5.5; a fortieth on Fable and Mythos 5.1 and 5.2",
          "1.25 times the input price for 5 minutes, twice for 1 hour",
          "512 tokens on the newest models; 1,024 to 4,096 on older ones",
          "5 minutes or 1 hour, renewed by each read",
          "No: one cache_control field turns it on",
          "Discounts stack; hits are best effort"
        ],
        [
          "OpenAI (GPT-5.6 and later)",
          "A tenth of the input price; a twentieth on GPT-6.1 Sol and Terra",
          "1.25 times the input price for 30 minutes, 2.5 times for 12 hours",
          "1,024 tokens",
          "At least 30 minutes, renewed by a read, or 12 hours",
          "Yes",
          "Price list shows cached input at half; the batch guide does not mention caching"
        ],
        [
          "Google (Gemini 4)",
          "A twentieth of the input price (a tenth on Gemini 3 models)",
          "No charge; explicit caches on Gemini 3 models pay storage by the hour",
          "4,096 tokens",
          "Not stated",
          "Yes",
          "Price list shows cached input at half; the batch guide is silent on implicit caching"
        ]
      ]
    },
    {
      "key": "cost-of-1-000-agent-calls-under-each-lever",
      "title": "Cost of 1,000 agent calls under each lever",
      "type": "chart",
      "source": "My arithmetic from Anthropic and OpenAI list prices, read 16 November 2026",
      "caption": "Each call: 10,000 input and 1,000 output tokens at $2 and $10 per million. Cache reads at $0.10 per million; all input that is not read from the cache is written to it at $2.50, the worst case. The routing bar assumes the cheaper model needs the same number of calls and succeeds as often, which is the thing to test.",
      "csv": "https://bardak.ai/data/prompt-caching-batching-routing/cost-of-1-000-agent-calls-under-each-lever.csv",
      "header": [
        "Setup",
        "Cost"
      ],
      "rows": [
        [
          "Caching on, nothing reused",
          "35.00"
        ],
        [
          "No discounts",
          "30.00"
        ],
        [
          "Caching, 50% of input read from cache",
          "23.00"
        ],
        [
          "Half the calls on a model at a twentieth of the price",
          "15.75"
        ],
        [
          "Batch",
          "15.00"
        ],
        [
          "Caching, 90% of input read from cache",
          "13.40"
        ],
        [
          "Batch and caching at 90%",
          "6.70"
        ]
      ]
    },
    {
      "key": "appraising-the-evidence",
      "title": "Appraising the evidence",
      "type": "benchmarkTable",
      "source": "",
      "caption": "",
      "csv": "https://bardak.ai/data/prompt-caching-batching-routing/appraising-the-evidence.csv",
      "header": [
        "Evidence",
        "Type",
        "Strength",
        "Main limitation"
      ],
      "rows": [
        [
          "Caching and batch documentation, price pages",
          "Vendor",
          "Authoritative for prices and rules",
          "Silent on typical hit rates; rules change monthly; Google's pages disagree on batch"
        ],
        [
          "Hit-rate figures (OpenAI examples, Anthropic range)",
          "Vendor",
          "The only published figures",
          "Examples, not measurements of typical use"
        ],
        [
          "RouteLLM, FrugalGPT",
          "Papers",
          "Peer reviewed, methods open",
          "Chat and classification, older models, partly LLM-judged, flattering baselines"
        ],
        [
          "RouterBench, LLMRouterBench, RouterArena",
          "Benchmarks",
          "Same questions for every router; mostly the same models",
          "One model per question or task, no routing inside an agent; RouterBench is by a router vendor"
        ],
        [
          "TwinRouterBench, LangChain test of NVIDIA's router",
          "Vendor benchmarks",
          "Real agent tasks, real spend",
          "One run or five, small, own product wins or is sold"
        ],
        [
          "My cost chart",
          "Own arithmetic",
          "Transparent, reproducible",
          "One call shape, list prices, worst-case writes"
        ]
      ]
    }
  ]
}