Add GPT-5.6 Luna BYOM pricing to the voice cost calculator

tools/cost_calculator/README.md

@@ -95,6 +95,15 @@ The GPT Live meter is hourly and is divided by 60 for USD-per-minute pricing.
   GPT Realtime 2.1 mini, GPT OSS 120B, and GPT-6 Sol are prefilled from the
   published South India meters. Global describes processing scope, not a
   promise of India-only inference.
+- **GPT-5.6 Luna BYOM** uses South India Global short-context Standard
+  inference prices: **$0.20 input / $0.02 cached input / $1.20 output per
+  1M tokens**, verified on 2026-10-26. These three meters record their own
+  retrieval date; the remaining snapshot retains its original date.
+  Luna is available for the text-LLM cascade and GPT Live backend scenarios,
+  with Central India BYO speech rates for the cascade. Managed-mode rows
+  use the existing explicitly labeled Global-model/BYO-speech reference,
+  not a verified managed Luna tariff. Long-context, Flex, Priority, and
+  cache-write rates are not substituted.
 - **GPT-6 reasoning is explicitly GPT-6 Sol**, using South India Global
   short-context Standard prices: **$2 input / $0.20 cached input / $10 output
   per 1M tokens**. The generic, unspecified GPT-6 placeholder is replaced

tools/cost_calculator/calculator.mjs

@@ -17,6 +17,7 @@ export const MODELS = [
   { id: "gpt-4.1", name: "GPT-4.1", tier: "Pro" },
   { id: "gpt-4.1-mini", name: "GPT-4.1 mini", tier: "Std" },
   { id: "gpt-4.1-nano", name: "GPT-4.1 nano", tier: "Lite" },
+  { id: "gpt-5.6-luna", name: "GPT-5.6 Luna" },
   { id: "gpt-5.4", name: "GPT-5.4", tier: "Pro", byomPrefix: "5.4" },
   { id: "gpt-5.4-mini", name: "GPT-5.4 mini", byomPrefix: "5.4 mini" },
   { id: "gpt-5.4-nano", name: "GPT-5.4 nano", byomPrefix: "5.4 nano" },
@@ -113,10 +114,11 @@ export function getRates(snapshot, model, hosting, pipeline = "cascade") {
       // pricing for reused tokens rather than inventing a cache discount.
       rates.textCached = rates.textInput;
       sources.textCached = sources.textInput;
-    } else if (model.id === "gpt-6-sol") {
-      read("textInput", "6-sol ShortCo Inp Std Gl 1M Tokens");
-      read("textCached", "6-sol ShortCo Cd Inp Std Gl 1M Tokens");
-      read("textOutput", "6-sol ShortCo Opt Std Gl 1M Tokens");
+    } else if (model.id === "gpt-6-sol" || model.id === "gpt-5.6-luna") {
+      const p = model.id === "gpt-6-sol" ? "6-sol" : "5.6 luna";
+      read("textInput", `${p} ShortCo Inp Std Gl 1M Tokens`);
+      read("textCached", `${p} ShortCo Cd Inp Std Gl 1M Tokens`);
+      read("textOutput", `${p} ShortCo Opt Std Gl 1M Tokens`);
     } else if (model.byomPrefix) {
       const p = model.byomPrefix;
       read("textInput", `${p} ${p.startsWith("5.4 ") ? "Inp" : "inp"} Gl 1M Tokens`);

tools/cost_calculator/calculator.test.mjs

@@ -55,7 +55,7 @@ test("South India GPT-4.1 and realtime prices do not reuse managed tier rates",
 });
 
 test("unverified managed tiers use explicit Global model references, not invented tier rates", () => {
-  for (const id of ["gpt-5.4-mini", "gpt-5.4-nano", "gpt-oss-120b", "gpt-6-sol"]) {
+  for (const id of ["gpt-5.6-luna", "gpt-5.4-mini", "gpt-5.4-nano", "gpt-oss-120b", "gpt-6-sol"]) {
     const reference = getRates(snapshot, model(id), "managed");
     assert.equal(reference.usesGlobalReference, true);
     assert.deepEqual(reference.rates, book(id, "byom"));
@@ -68,6 +68,27 @@ test("unverified managed tiers use explicit Global model references, not invente
   close(tier.rates.textInput, 4);
 });
 
+test("GPT-5.6 Luna BYOM uses exact Standard Global inference meters and BYO speech", () => {
+  const { rates, sources } = getRates(snapshot, model("gpt-5.6-luna"), "byom");
+  close(rates.textInput, 0.2);
+  close(rates.textCached, 0.02);
+  close(rates.textOutput, 1.2);
+  close(rates.speechInput, 12.5);
+  close(rates.speechOutput, 23);
+  for (const key of ["textInput", "textCached", "textOutput"]) {
+    assert.match(sources[key].meterName, /^5\.6 luna ShortCo .* Std Gl 1M Tokens$/);
+    assert.equal(sources[key].retrievedAt, "2026-10-26");
+    assert.equal(sources[key].armRegionName, "southindia");
+    assert.equal(sources[key].processingScope, "Global");
+  }
+  const textCost = (5000 * 0.2 + 15000 * 0.02 + 150 * 0.2 + 450 * 0.02 + 150 * 1.2) / 1e6;
+  const cascade = calculate(DEFAULTS, "cascade", rates);
+  close(cascade.list, textCost + (360 * 12.5 + 720 * 23) / 1e6);
+  close(cascade.listPerMinute, cascade.list / 1.5);
+  const live = calculate(DEFAULTS, "live", book("gpt-5.6-luna", "byom", "live"));
+  close(live.list, textCost + 1.5 * 0.05);
+});
+
 test("all selectable models have priced scenarios across both hosting modes", () => {
   for (const hosting of ["managed", "byom"]) for (const m of MODELS) {
     const pipeline = m.realtime ? "native" : "cascade";

tools/cost_calculator/prices.json

@@ -9,6 +9,42 @@
     "armRegionName eq 'southindia' and serviceName eq 'Foundry Models' and priceType eq 'Consumption'"
   ],
   "meters": [
+    {
+      "productName": "Azure OpenAI GPT5",
+      "skuName": "5.6 luna ShortCo Inp Std Gl",
+      "meterName": "5.6 luna ShortCo Inp Std Gl 1M Tokens",
+      "armRegionName": "southindia",
+      "processingScope": "Global",
+      "retailPrice": 0.2,
+      "unitOfMeasure": "1M",
+      "meterId": "a05ffb45-17d1-5a56-a43e-f8af27b7e7e5",
+      "effectiveStartDate": "2026-08-01T00:00:00Z",
+      "retrievedAt": "2026-10-26"
+    },
+    {
+      "productName": "Azure OpenAI GPT5",
+      "skuName": "5.6 luna ShortCo Cd Inp Std Gl",
+      "meterName": "5.6 luna ShortCo Cd Inp Std Gl 1M Tokens",
+      "armRegionName": "southindia",
+      "processingScope": "Global",
+      "retailPrice": 0.02,
+      "unitOfMeasure": "1M",
+      "meterId": "5489ae4f-17c9-5271-931c-da4e2289dffc",
+      "effectiveStartDate": "2026-08-01T00:00:00Z",
+      "retrievedAt": "2026-10-26"
+    },
+    {
+      "productName": "Azure OpenAI GPT5",
+      "skuName": "5.6 luna ShortCo Opt Std Gl",
+      "meterName": "5.6 luna ShortCo Opt Std Gl 1M Tokens",
+      "armRegionName": "southindia",
+      "processingScope": "Global",
+      "retailPrice": 1.2,
+      "unitOfMeasure": "1M",
+      "meterId": "d01c83fd-4a6c-5c28-b962-9e3de99eb4cd",
+      "effectiveStartDate": "2026-08-01T00:00:00Z",
+      "retrievedAt": "2026-10-26"
+    },
     {
       "productName": "Azure Speech",
       "skuName": null,