{
  "schema": "s2s-bench-v1",
  "kind": "bill",
  "round": "opencall-addendum-qwen38",
  "currency": "USD",
  "what_this_is": "the post-publication addendum round's bill. It is a SEPARATE round from the sealed twenty and a separate bill: bill.json covers opencall-r2 and is unchanged by this file.",
  "figures_are": "upper bounds by construction: actual token counts from each endpoint's own counters, priced at the UNCACHED input rate",
  "how_every_line_re_derives": "each receipt carries the token counts it was billed on and the rate they were priced at, and `usd_re_derived_from_these_tokens` is that multiplication performed: prompt × input/1e6 + completion × output/1e6. Sum the receipts to get the line, sum the lines to get the total. The builder asserts all three before it will write this file, and asserts that the total is the $0.3579 the exhibit page prints.",
  "why_this_file_exists": "the addendum published on 2026-08-16 with $0.3579 on the page and no token counts behind it anywhere in this kit — the one figure on that page a reader could not re-derive. A content-accuracy pass marked it as such the same day, and this file, shipped the same evening, is what removes the mark. Every count in it was recorded when the round ran: this is a join, never a re-run, and no model was called to produce it.",
  "lines": [
    {
      "line": "openai-flagship",
      "model": "gpt-5.5-2026-04-23",
      "family": "openai",
      "transport": "openai-api",
      "billed_by": "the addendum round's own judge run, results/judge/opencall-addendum-qwen38/runs/RUN-openai-flagship.json — one file per seat, and the whole of that seat's spend on this round",
      "calls": 4,
      "prompt_tokens": 21903,
      "completion_tokens": 4947,
      "pricing_usd_per_mtok": {
        "input": 5.0,
        "cached_input": 0.5,
        "output": 30.0
      },
      "usd_upper_bound": 0.2579,
      "usd_re_derived_from_these_tokens": 0.257925,
      "receipts": [
        {
          "call": "house-fable|S1-ask-A",
          "prompt_tokens": 5389,
          "completion_tokens": 1316,
          "pricing_usd_per_mtok": {
            "input": 5.0,
            "cached_input": 0.5,
            "output": 30.0
          },
          "usd": 0.066425,
          "usd_re_derived_from_these_tokens": 0.066425
        },
        {
          "call": "house-fable|S1-ask-B",
          "prompt_tokens": 5592,
          "completion_tokens": 1793,
          "pricing_usd_per_mtok": {
            "input": 5.0,
            "cached_input": 0.5,
            "output": 30.0
          },
          "usd": 0.08175,
          "usd_re_derived_from_these_tokens": 0.08175
        },
        {
          "call": "house-fable|S2-ask-A",
          "prompt_tokens": 5435,
          "completion_tokens": 1127,
          "pricing_usd_per_mtok": {
            "input": 5.0,
            "cached_input": 0.5,
            "output": 30.0
          },
          "usd": 0.060985,
          "usd_re_derived_from_these_tokens": 0.060985
        },
        {
          "call": "house-fable|S3-ask-A",
          "prompt_tokens": 5487,
          "completion_tokens": 711,
          "pricing_usd_per_mtok": {
            "input": 5.0,
            "cached_input": 0.5,
            "output": 30.0
          },
          "usd": 0.048765,
          "usd_re_derived_from_these_tokens": 0.048765
        }
      ],
      "cap_usd": 20.0,
      "cap_basis": "prereg §3.4: OpenAI <= $20.00 across the key, from the cited rates against measured tokens.",
      "cost_state": "figure"
    },
    {
      "line": "kimi-k3",
      "model": "kimi-k3",
      "family": "moonshot",
      "transport": "ollama-cloud",
      "billed_by": "the addendum round's own judge run, results/judge/opencall-addendum-qwen38/runs/RUN-kimi-k3.json — one file per seat, and the whole of that seat's spend on this round",
      "calls": 5,
      "prompt_tokens": 27899,
      "completion_tokens": 1084,
      "pricing_usd_per_mtok": {
        "input": 3.0,
        "cached_input": 0.3,
        "output": 15.0
      },
      "usd_upper_bound": 0.1,
      "usd_re_derived_from_these_tokens": 0.099957,
      "receipts": [
        {
          "call": "house-fable|S1-ask-A",
          "prompt_tokens": 5443,
          "completion_tokens": 264,
          "pricing_usd_per_mtok": {
            "input": 3.0,
            "cached_input": 0.3,
            "output": 15.0
          },
          "usd": 0.020289,
          "usd_re_derived_from_these_tokens": 0.020289
        },
        {
          "call": "house-fable|S1-ask-B",
          "prompt_tokens": 5641,
          "completion_tokens": 280,
          "pricing_usd_per_mtok": {
            "input": 3.0,
            "cached_input": 0.3,
            "output": 15.0
          },
          "usd": 0.021123,
          "usd_re_derived_from_these_tokens": 0.021123
        },
        {
          "call": "house-fable|S2-ask-A",
          "prompt_tokens": 5484,
          "completion_tokens": 193,
          "pricing_usd_per_mtok": {
            "input": 3.0,
            "cached_input": 0.3,
            "output": 15.0
          },
          "usd": 0.019347,
          "usd_re_derived_from_these_tokens": 0.019347
        },
        {
          "call": "house-fable|S2-ask-A | format-reminder",
          "prompt_tokens": 5796,
          "completion_tokens": 177,
          "pricing_usd_per_mtok": {
            "input": 3.0,
            "cached_input": 0.3,
            "output": 15.0
          },
          "usd": 0.020043,
          "usd_re_derived_from_these_tokens": 0.020043
        },
        {
          "call": "house-fable|S3-ask-A",
          "prompt_tokens": 5535,
          "completion_tokens": 170,
          "pricing_usd_per_mtok": {
            "input": 3.0,
            "cached_input": 0.3,
            "output": 15.0
          },
          "usd": 0.019155,
          "usd_re_derived_from_these_tokens": 0.019155
        }
      ],
      "cap_usd": 18.49,
      "cap_basis": "prereg §3.4: kimi-k3 <= $18.49, stated as DERIVED headroom, with the balance probe's $0.000363 already on the bill.",
      "cost_state": "figure"
    }
  ],
  "seats_that_bill_nothing": [
    {
      "line": "deepseek-v4-pro",
      "model": "deepseek-v4-pro:preview",
      "family": "deepseek",
      "transport": "ollama-cloud",
      "billed_by": "the addendum round's own judge run, results/judge/opencall-addendum-qwen38/runs/RUN-deepseek-v4-pro.json — one file per seat, and the whole of that seat's spend on this round",
      "calls": 4,
      "prompt_tokens": 22140,
      "completion_tokens": 978,
      "pricing_usd_per_mtok": {
        "input": 0.0,
        "cached_input": 0.0,
        "output": 0.0
      },
      "usd_upper_bound": 0.0,
      "usd_re_derived_from_these_tokens": 0.0,
      "receipts": [
        {
          "call": "house-opus|S1-ask-A",
          "prompt_tokens": 5453,
          "completion_tokens": 309,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S1-ask-B",
          "prompt_tokens": 5645,
          "completion_tokens": 282,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S2-ask-A",
          "prompt_tokens": 5494,
          "completion_tokens": 190,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S3-ask-A",
          "prompt_tokens": 5548,
          "completion_tokens": 197,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        }
      ],
      "cost_state": "$0.00*",
      "basis": "actual token counts from the endpoint's own prompt_eval_count / eval_count, priced at the UNCACHED input rate. Cached input bills at a tenth and is not reported by the API, so this is an upper bound, never an under-count."
    },
    {
      "line": "mistral-large-3-675b",
      "model": "mistral-large-3:675b",
      "family": "mistral",
      "transport": "ollama-cloud",
      "billed_by": "the addendum round's own judge run, results/judge/opencall-addendum-qwen38/runs/RUN-mistral-large-3-675b.json — one file per seat, and the whole of that seat's spend on this round",
      "calls": 4,
      "prompt_tokens": 23052,
      "completion_tokens": 977,
      "pricing_usd_per_mtok": {
        "input": 0.0,
        "cached_input": 0.0,
        "output": 0.0
      },
      "usd_upper_bound": 0.0,
      "usd_re_derived_from_these_tokens": 0.0,
      "receipts": [
        {
          "call": "house-opus|S1-ask-A",
          "prompt_tokens": 5683,
          "completion_tokens": 300,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S1-ask-B",
          "prompt_tokens": 5876,
          "completion_tokens": 278,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S2-ask-A",
          "prompt_tokens": 5713,
          "completion_tokens": 197,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        },
        {
          "call": "house-opus|S3-ask-A",
          "prompt_tokens": 5780,
          "completion_tokens": 202,
          "pricing_usd_per_mtok": {
            "input": 0.0,
            "cached_input": 0.0,
            "output": 0.0
          },
          "usd": 0.0,
          "usd_re_derived_from_these_tokens": 0.0
        }
      ],
      "cost_state": "$0.00*",
      "basis": "actual token counts from the endpoint's own prompt_eval_count / eval_count, priced at the UNCACHED input rate. Cached input bills at a tenth and is not reported by the API, so this is an upper bound, never an under-count."
    },
    {
      "line": "gemma4-26b",
      "model": "gemma4:26b",
      "family": "google",
      "transport": "local-ollama",
      "billed_by": "the addendum round's own judge run, results/judge/opencall-addendum-qwen38/runs/RUN-gemma4-26b.json — one file per seat, and the whole of that seat's spend on this round",
      "cost_state": "$0.00 local",
      "basis": "this seat bills nothing: a local daemon on a box we own.",
      "prompt_tokens": null,
      "completion_tokens": null
    }
  ],
  "metered_seats": 2,
  "metered_calls": 9,
  "metered_total_usd_upper_bound": 0.3579,
  "metered_total_usd_re_derived": 0.357882,
  "totals_note": "the two totals agree at four decimal places (0.3579). They are computed differently on purpose — one sums each line's own already-rounded figure, the other sums the unrounded products — and both print, so a round where they parted in the last place would show it rather than hide it.",
  "relationship_to_bill_json": {
    "sealed_round_metered_total_usd": 2.2826,
    "this_round_metered_total_usd": 0.3579,
    "exhibit_total_usd": 2.6405,
    "note": "these are ADDITIVE, not overlapping: no call is on both bills. The exhibit page prints them separately, because the sealed round's figure is a receipt for a round that is frozen and this one is a receipt for a round that came after it. The sum is published here so a reader who wants the whole exhibit's cost does not have to guess whether adding them double-counts. It does not."
  },
  "cost_states": {
    "figure": "metered, priced at a cited rate against measured tokens",
    "$0.00*": "plan-included: no per-token charge on this key, but a real, unitemisable subscription sits behind the zero. The token counts are published anyway, because they are what a price would be applied to if one existed.",
    "$0.00 local": "answered from a local daemon, where no bill exists to read. Its token counts read null rather than zero — the round holds none for this seat, and a zero would claim it did."
  },
  "not_on_this_bill": "the addendum arm itself (qwen3.8:27b) ran on a local daemon and cost nothing to meter, and its three discarded cold-start warmups are recorded in addendum-samples.json rather than here — there is no dollar for them to carry."
}