slash-tokens 1.6.5 → 1.6.6

package/CHANGELOG.md

@@ -1,5 +1,28 @@
 # Changelog
 
+## [1.6.6] — The Fixed Deal Edition
+
+*2026-10-07*
+
+`--version` / `--help` answer, and today's models price correctly.
+
+Solo $20 mailbox, 10% waived. Team $39 for the data.
+
+### Fixed
+- **`slash-tokens --version` and `--help` print and exit.** In 1.6.5 every flag fell through to a full scan of the current folder, so `--version` run from `~` scanned the home folder. An empty `bunx` run now tells you to run it in an app.
+- **`preflight()` no longer throws on current models or real API IDs.** IDs are canonicalised strictly (`claude-opus-4-7` → `claude-opus-4.7`, date stamps dropped). There is no family guessing: an unknown version still fails with "Unknown model", so a new model never silently gets an older model's price.
+- **`npm publish` always builds first** (`prepublishOnly`), so a stale `dist/` can't ship again (1.6.4).
+- **`report()` sends numbers only.** A test pins its payload to `tokens_estimated`, `tokens_saved`, `model`, `action`, `cost_saved_usd`; prompt content never leaves the machine.
+- **`npm test` runs offline.** The live integration suite (`tests/z-integration.test.ts`, which registers keys on mcpaas.live) runs only with `SLASH_LIVE=1`; the weekly `integration.yml` sets it.
+
+### Changed
+- **Model table, prices as of 2026-10-07** (checked against the Anthropic, OpenAI, xAI and Google pricing pages). Added Claude Opus 5.5 ($4/$20), Sonnet 5.5 ($2/$10), Opus 4.8 / 4.6 / 4.5, Sonnet 4.6 / 4.5, Fable 5 / 5.1 and Mythos 5 / 5.1 ($10/$50); Grok 4.7 and 4.5 ($2/$6, doubling above 200K), grok-build-0.1 ($1/$2); GPT-6 Astra ($10/$50), GPT-6.1 Sol and GPT-6 Sol ($2/$10), GPT-6 Luna ($0.10/$0.50); Gemini 3.8 / 3.7 / 3.6 Flash ($0.75/$3.75 launch price through 2026-12-31), 3.5 Flash ($1.50/$9), 3.1 Flash-Lite ($0.25/$1.50), 3.1 Pro Preview.
+- **`slash-tokens/auto` routing is unchanged within 1.6.x.** `/auto` rewrites live requests only to the models it routed to in 1.6.5 (`AUTO_ROUTE_TARGETS`), using the same identification, so a patch upgrade never changes where production calls go. New models are now recognised and priced correctly in `/auto` events (`identifyModel`: `claude-opus-5-5` at $4/$20, `gpt-6-astra` at $10/$50); /auto starts routing to them in 1.7.0.
+- **Routing (`preflightRoute`) follows the new ladder.** Cheapest same-provider is now GPT-6 Luna for OpenAI and Gemini 3.1 Flash-Lite for Google (Anthropic and xAI unchanged). grok-build-0.1 is priced but never a routing target (`NOT_ROUTE_TARGETS`): it's a coding-agent model. When two alternatives' costs round to the same value (tiny prompts), the lower list price wins.
+- **Calibration:** new Claude, Grok and Gemini models use their family's measured factor. GPT-6 is a new generation with an unbenchmarked tokenizer, so it takes the conservative default until a bench run adds it (slash never under-reports).
+
+No Team/Solo price change.
+
 ## [1.6.5] — The Fixed Deal Edition
 
 *2026-08-25*

package/README.md

@@ -15,10 +15,14 @@ Know the cost before the call leaves your machine.
 Models change. Windows grow. Slash adapts — you keep building.
 Cheaper tokens haven't shrunk the bill — usage has.
 
-## v1.6.5 — The Fixed Deal Edition
+## v1.6.6 — The Fixed Deal Edition
+
+`--version` / `--help` answer, and today's models price correctly.
 
 Solo $20 mailbox, 10% waived. Team $39 for the data.
 
+New in 1.6.6: Claude Opus 5.5 / Sonnet 5.5 / Fable 5.1, Grok 4.7, GPT-6 (Astra, Sol, Luna), Gemini 3.6–3.8 Flash and 3.1 Flash-Lite, priced as of 2026-10-07. Real API IDs (`claude-opus-4-7`) work in `preflight()`. `preflightRoute()` now finds GPT-6 Luna and Gemini 3.1 Flash-Lite as the cheapest same-provider options.
+
 **Free forever is bunx** — no account. A one-person account is email → key, **$20 on the house**. We show the savings. We don't charge. 10% is the model, waived. Team is **$39 for the data** (`$390`/year).
 
 ```bash
@@ -26,7 +30,7 @@ bunx slash-tokens
 # or: npx --yes slash-tokens
 ```
 
-Run it in a project that already calls an LLM. An empty folder prints that nothing was found — that's normal.
+Run it in a project that already calls an LLM. An empty folder prints that nothing was found, then tells you to run it in an app — that's normal. `--version` / `--help` print and exit (they do not scan). Pin proof: `slash-tokens --version` or `npm view slash-tokens version`.
 
 ```bash
 npm install slash-tokens

package/dist/cli.js

@@ -1,5 +1,10 @@
 #!/usr/bin/env node
 
+// src/cli.ts
+import { readFileSync as readFileSync2 } from "node:fs";
+import { dirname, join as join2 } from "node:path";
+import { fileURLToPath } from "node:url";
+
 // src/scanner.ts
 import { readdirSync, readFileSync, statSync } from "fs";
 import { join, extname } from "path";
@@ -49,10 +54,137 @@ function writeToMemory(content) {
   return maxLen;
 }
 
+// src/models.ts
+var OPUS = { input: 5, output: 25, context: 1e6 };
+var OPUS_55 = { input: 4, output: 20, context: 1e6 };
+var FABLE = { input: 10, output: 50, context: 1e6 };
+var SONNET_4X = { input: 3, output: 15, context: 1e6 };
+var SONNET = { input: 2, output: 10, context: 1e6 };
+var HAIKU = { input: 1, output: 5, context: 200000 };
+var GROK_46 = {
+  input: 2,
+  output: 6,
+  context: 500000,
+  longContextThreshold: 200000,
+  longContextInput: 4,
+  longContextOutput: 12
+};
+var GROK_43 = {
+  input: 1.25,
+  output: 2.5,
+  context: 1e6,
+  longContextThreshold: 200000,
+  longContextInput: 2.5,
+  longContextOutput: 5
+};
+var GEMINI_PRO = {
+  input: 2,
+  output: 12,
+  context: 1e6,
+  longContextThreshold: 200000,
+  longContextInput: 4,
+  longContextOutput: 18
+};
+var GEMINI_FLASH = { input: 0.3, output: 2.5, context: 1e6 };
+var GEMINI_FLASH_3X = { input: 0.75, output: 3.75, context: 1e6 };
+var GEMINI_35_FLASH = { input: 1.5, output: 9, context: 1e6 };
+var GEMINI_31_FLASH_LITE = { input: 0.25, output: 1.5, context: 1e6 };
+var GROK_BUILD = {
+  input: 1,
+  output: 2,
+  context: 256000,
+  longContextThreshold: 200000,
+  longContextInput: 2,
+  longContextOutput: 4
+};
+var GPT_6_ASTRA = { input: 10, output: 50, context: 1050000 };
+var GPT_6_SOL = { input: 2, output: 10, context: 1050000 };
+var GPT_6_LUNA = { input: 0.1, output: 0.5, context: 1050000 };
+var GPT_SOL = { input: 4, output: 20, context: 1050000 };
+var GPT_TERRA = { input: 2, output: 12, context: 1050000 };
+var GPT_LUNA = { input: 0.2, output: 1.2, context: 1050000 };
+var GPT_54 = { input: 2.5, output: 15, context: 1e6 };
+var GPT_54_MINI = { input: 0.75, output: 4.5, context: 128000 };
+var GPT_54_NANO = { input: 0.2, output: 1.25, context: 128000 };
+var MODELS = {
+  "claude-fable-5.1": { ...FABLE },
+  "claude-fable-5": { ...FABLE },
+  "claude-mythos-5.1": { ...FABLE },
+  "claude-mythos-5": { ...FABLE },
+  "claude-opus-5.5": { ...OPUS_55 },
+  "claude-opus-5": { ...OPUS },
+  "claude-opus-4.8": { ...OPUS },
+  "claude-opus": { ...OPUS },
+  "claude-opus-4.7": { ...OPUS },
+  "claude-opus-4.6": { ...OPUS },
+  "claude-opus-4.5": { ...OPUS },
+  "claude-sonnet-5.5": { ...SONNET },
+  "claude-sonnet-5": { ...SONNET },
+  "claude-sonnet": { ...SONNET },
+  "claude-sonnet-4.6": { ...SONNET_4X },
+  "claude-sonnet-4.5": { ...SONNET_4X },
+  "claude-haiku-4.5": { ...HAIKU },
+  "claude-haiku": { ...HAIKU },
+  "grok-4.7": { ...GROK_46 },
+  "grok-4.6": { ...GROK_46 },
+  "grok-4.5": { ...GROK_46 },
+  "grok-build-0.1": { ...GROK_BUILD },
+  "grok-4.3": { ...GROK_43 },
+  "grok-4.20": { ...GROK_43 },
+  "grok-4-1-fast": { ...GROK_43 },
+  "gemini-3.1-pro": { ...GEMINI_PRO },
+  "gemini-3.1-pro-preview": { ...GEMINI_PRO },
+  "gemini-3.8-flash": { ...GEMINI_FLASH_3X },
+  "gemini-3.7-flash": { ...GEMINI_FLASH_3X },
+  "gemini-3.6-flash": { ...GEMINI_FLASH_3X },
+  "gemini-3.5-flash": { ...GEMINI_35_FLASH },
+  "gemini-3.1-flash-lite": { ...GEMINI_31_FLASH_LITE },
+  "gemini-3.5-flash-lite": { ...GEMINI_FLASH },
+  "gemini-2.5-flash": { ...GEMINI_FLASH },
+  "gpt-6-astra": { ...GPT_6_ASTRA },
+  "gpt-6.1-sol": { ...GPT_6_SOL },
+  "gpt-6-sol": { ...GPT_6_SOL },
+  "gpt-6-luna": { ...GPT_6_LUNA },
+  "gpt-5.6-sol": { ...GPT_SOL },
+  "gpt-5.6-terra": { ...GPT_TERRA },
+  "gpt-5.6-luna": { ...GPT_LUNA },
+  "gpt-5.4": { ...GPT_54 },
+  "gpt-5.4-mini": { ...GPT_54_MINI },
+  "gpt-5.4-nano": { ...GPT_54_NANO }
+};
+function canonicalModel(name) {
+  if (MODELS[name])
+    return name;
+  let n = name.trim().toLowerCase();
+  if (MODELS[n])
+    return n;
+  n = n.replace(/-\d{8}$/, "");
+  if (MODELS[n])
+    return n;
+  const dotted = n.replace(/(\d)-(\d)(?=$|-)/g, "$1.$2");
+  if (MODELS[dotted])
+    return dotted;
+  return n;
+}
+function getModel(name) {
+  return MODELS[canonicalModel(name)];
+}
+
 // src/slash.ts
 var WASM_INPUT_OFFSET2 = 4096;
 var CALIBRATION = {
+  "claude-fable-5.1": 2.05,
+  "claude-fable-5": 2.05,
+  "claude-mythos-5.1": 2.05,
+  "claude-mythos-5": 2.05,
+  "claude-opus-5.5": 2.05,
   "claude-opus-5": 2.05,
+  "claude-opus-4.8": 2.05,
+  "claude-opus-4.6": 2.05,
+  "claude-opus-4.5": 2.05,
+  "claude-sonnet-5.5": 2.05,
+  "claude-sonnet-4.6": 2.05,
+  "claude-sonnet-4.5": 2.05,
   "claude-opus": 2.05,
   "claude-opus-4.7": 2.05,
   "claude-sonnet-5": 2.05,
@@ -60,9 +192,18 @@ var CALIBRATION = {
   "claude-haiku-4.5": 1.45,
   "claude-haiku": 1.45,
   "gemini-3.1-pro": 1.45,
+  "gemini-3.1-pro-preview": 1.45,
+  "gemini-3.8-flash": 1.45,
+  "gemini-3.7-flash": 1.45,
+  "gemini-3.6-flash": 1.45,
+  "gemini-3.5-flash": 1.45,
+  "gemini-3.1-flash-lite": 1.45,
   "gemini-3.5-flash-lite": 1.45,
   "gemini-2.5-flash": 1.45,
+  "grok-4.7": 1.15,
   "grok-4.6": 1.15,
+  "grok-4.5": 1.15,
+  "grok-build-0.1": 1.15,
   "grok-4.3": 1.15,
   "grok-4.20": 1.15,
   "grok-4-1-fast": 1.15,
@@ -82,7 +223,7 @@ function slash(content, model) {
   const raw = instance.exports.estimate_tokens(WASM_INPUT_OFFSET2, len);
   if (!model)
     return raw;
-  const factor = CALIBRATION[model] ?? DEFAULT_UNKNOWN_MODEL_FACTOR;
+  const factor = CALIBRATION[canonicalModel(model)] ?? DEFAULT_UNKNOWN_MODEL_FACTOR;
   return factor === 1 ? raw : Math.ceil(raw * factor);
 }
 
@@ -211,67 +352,6 @@ function scan(dir) {
   };
 }
 
-// src/models.ts
-var OPUS = { input: 5, output: 25, context: 1e6 };
-var SONNET = { input: 2, output: 10, context: 1e6 };
-var HAIKU = { input: 1, output: 5, context: 200000 };
-var GROK_46 = {
-  input: 2,
-  output: 6,
-  context: 500000,
-  longContextThreshold: 200000,
-  longContextInput: 4,
-  longContextOutput: 12
-};
-var GROK_43 = {
-  input: 1.25,
-  output: 2.5,
-  context: 1e6,
-  longContextThreshold: 200000,
-  longContextInput: 2.5,
-  longContextOutput: 5
-};
-var GEMINI_PRO = {
-  input: 2,
-  output: 12,
-  context: 1e6,
-  longContextThreshold: 200000,
-  longContextInput: 4,
-  longContextOutput: 18
-};
-var GEMINI_FLASH = { input: 0.3, output: 2.5, context: 1e6 };
-var GPT_SOL = { input: 4, output: 20, context: 1050000 };
-var GPT_TERRA = { input: 2, output: 12, context: 1050000 };
-var GPT_LUNA = { input: 0.2, output: 1.2, context: 1050000 };
-var GPT_54 = { input: 2.5, output: 15, context: 1e6 };
-var GPT_54_MINI = { input: 0.75, output: 4.5, context: 128000 };
-var GPT_54_NANO = { input: 0.2, output: 1.25, context: 128000 };
-var MODELS = {
-  "claude-opus-5": { ...OPUS },
-  "claude-opus": { ...OPUS },
-  "claude-opus-4.7": { ...OPUS },
-  "claude-sonnet-5": { ...SONNET },
-  "claude-sonnet": { ...SONNET },
-  "claude-haiku-4.5": { ...HAIKU },
-  "claude-haiku": { ...HAIKU },
-  "grok-4.6": { ...GROK_46 },
-  "grok-4.3": { ...GROK_43 },
-  "grok-4.20": { ...GROK_43 },
-  "grok-4-1-fast": { ...GROK_43 },
-  "gemini-3.1-pro": { ...GEMINI_PRO },
-  "gemini-3.5-flash-lite": { ...GEMINI_FLASH },
-  "gemini-2.5-flash": { ...GEMINI_FLASH },
-  "gpt-5.6-sol": { ...GPT_SOL },
-  "gpt-5.6-terra": { ...GPT_TERRA },
-  "gpt-5.6-luna": { ...GPT_LUNA },
-  "gpt-5.4": { ...GPT_54 },
-  "gpt-5.4-mini": { ...GPT_54_MINI },
-  "gpt-5.4-nano": { ...GPT_54_NANO }
-};
-function getModel(name) {
-  return MODELS[name] || MODELS[name.toLowerCase()];
-}
-
 // src/report.ts
 var R = "\x1B[0m";
 var B = "\x1B[1m";
@@ -296,6 +376,7 @@ function printReport(sites, filesScanned, timeMs, cwd) {
   if (sites.length === 0) {
     console.log(`${WHITE}  No AI API call sites detected.${R}`);
     console.log(`${GRAY}  Supported: OpenAI, Anthropic, Vercel AI, LangChain, Gemini, Bedrock, Grok${R}`);
+    console.log(`${GRAY}  Run this in a project that already calls an LLM.${R}`);
     console.log("");
     return;
   }
@@ -367,6 +448,14 @@ function init(opts) {
 
 // src/cli.ts
 var args = process.argv.slice(2);
+if (args.includes("--version") || args.includes("-V")) {
+  console.log(packageVersion());
+  process.exit(0);
+}
+if (args.includes("--help") || args.includes("-h")) {
+  printHelp();
+  process.exit(0);
+}
 var keyArg = args.find((a) => a.startsWith("--key="));
 var key = keyArg?.split("=")[1] || process.env.SLASH_KEY;
 if (key)
@@ -388,3 +477,21 @@ if (sites.length > 0) {
   }
   console.log("");
 }
+function packageVersion() {
+  const here = dirname(fileURLToPath(import.meta.url));
+  const pkg = JSON.parse(readFileSync2(join2(here, "..", "package.json"), "utf8"));
+  return pkg.version;
+}
+function printHelp() {
+  console.log(`slash-tokens — Token Optimization for Context Engineers
+
+  bunx slash-tokens          try (scan this directory, no account)
+  npm install slash-tokens   SDK
+
+  --version, -V    print version and exit
+  --help, -h       print this help and exit
+  --key=KEY        optional; CLI scans never charge
+
+Run in a project that already calls an LLM.
+Empty folder → no call sites. That's normal.`);
+}

package/dist/intercept.d.ts

@@ -11,6 +11,14 @@ export interface InterceptEvent {
     routed: boolean;
     timestamp: string;
 }
+/**
+ * The model a request actually names, for PRICING: the exact table entry when
+ * the strict canonical ID matches (claude-opus-5-5 → claude-opus-5.5,
+ * gpt-6-luna), else the legacy family mapping. Routing keeps using
+ * normalizeModel() so /auto decisions are unchanged within 1.6.x.
+ */
+export declare function identifyModel(raw: string): string;
 export declare function normalizeModel(raw: string): string;
+export declare function findCheapestRoute(provider: string, tokens: number, currentModel: string): string | null;
 export declare function onIntercept(handler: (event: InterceptEvent) => void): void;
 export declare function patchFetch(): void;

package/dist/intercept.js

@@ -1,7 +1,7 @@
 import { slash } from './slash.js';
-import { getModel } from './models.js';
+import { getModel, MODELS, canonicalModel } from './models.js';
 import { shouldRoute, isModelAllowed } from './config.js';
-import { PROVIDER_MODELS } from './providers.js';
+import { PROVIDER_MODELS, AUTO_ROUTE_TARGETS, NOT_ROUTE_TARGETS } from './providers.js';
 // Reverse lookup: model name → provider model names in the API
 // (what to put back in the request body)
 const MODEL_API_NAMES = {
@@ -62,6 +62,16 @@ const AI_ENDPOINTS = [
 // (getModel() returning undefined → $0 reported cost, see slash.ts's
 // DEFAULT_UNKNOWN_MODEL_FACTOR comment for why "unrecognized" defaulting
 // to a falsely-safe-looking value is the dangerous case).
+/**
+ * The model a request actually names, for PRICING: the exact table entry when
+ * the strict canonical ID matches (claude-opus-5-5 → claude-opus-5.5,
+ * gpt-6-luna), else the legacy family mapping. Routing keeps using
+ * normalizeModel() so /auto decisions are unchanged within 1.6.x.
+ */
+export function identifyModel(raw) {
+    const c = canonicalModel(raw);
+    return MODELS[c] ? c : normalizeModel(raw);
+}
 export function normalizeModel(raw) {
     const lower = raw.toLowerCase();
     // Anthropic — specific versions before generic family
@@ -148,7 +158,7 @@ function extractContent(body) {
     return JSON.stringify(body);
 }
 // Find cheapest model from same provider that fits
-function findCheapestRoute(provider, tokens, currentModel) {
+export function findCheapestRoute(provider, tokens, currentModel) {
     const providerModels = PROVIDER_MODELS[provider];
     if (!providerModels)
         return null;
@@ -156,6 +166,10 @@ function findCheapestRoute(provider, tokens, currentModel) {
     for (const model of providerModels) {
         if (model === currentModel)
             continue;
+        if (!AUTO_ROUTE_TARGETS.has(model))
+            continue; // 1.6.x: /auto rewrites only to its 1.6.5 targets
+        if (NOT_ROUTE_TARGETS.has(model))
+            continue; // specialised models are never targets
         if (!isModelAllowed(model))
             continue; // user excluded this model
         const info = getModel(model);
@@ -188,13 +202,14 @@ export function patchFetch() {
                     const body = JSON.parse(bodyStr);
                     const content = extractContent(body);
                     const rawModel = match.modelExtractor(body, url);
-                    const originalModel = normalizeModel(rawModel);
+                    const routingModel = normalizeModel(rawModel); // decides routes (unchanged in 1.6.x)
+                    const originalModel = identifyModel(rawModel); // prices the request it really names
                     const tokens = slash(content, originalModel);
                     const originalInfo = getModel(originalModel);
                     const originalCost = originalInfo ? Math.round(((tokens / 1000000) * originalInfo.input) * 1000000) / 1000000 : 0;
                     const fits = originalInfo ? tokens <= originalInfo.context : true;
                     // Find cheapest route within same provider (if routing enabled)
-                    const routeModel = shouldRoute() ? findCheapestRoute(match.provider, tokens, originalModel) : null;
+                    const routeModel = shouldRoute() ? findCheapestRoute(match.provider, tokens, routingModel) : null;
                     const routedInfo = routeModel ? getModel(routeModel) : null;
                     const routedCost = routedInfo ? Math.round(((tokens / 1000000) * routedInfo.input) * 1000000) / 1000000 : originalCost;
                     const salvaged = routeModel ? Math.round((originalCost - routedCost) * 1000000) / 1000000 : 0;

package/dist/models.d.ts

@@ -7,6 +7,13 @@ export interface ModelInfo {
     longContextOutput?: number;
 }
 export declare const MODELS: Record<string, ModelInfo>;
+/**
+ * Real API IDs → table keys, strictly: lowercase, drop a trailing date stamp
+ * (`-20250514`), and write version numbers with dots (`claude-opus-4-7` →
+ * `claude-opus-4.7`). No family guessing: an unknown version stays unknown,
+ * so a new model never silently gets an older model's price.
+ */
+export declare function canonicalModel(name: string): string;
 export declare function getModel(name: string): ModelInfo | undefined;
 export declare function effectiveRate(tokens: number, info: ModelInfo): {
     input: number;

package/dist/models.js

@@ -1,4 +1,7 @@
 const OPUS = { input: 5.00, output: 25.00, context: 1000000 };
+const OPUS_55 = { input: 4.00, output: 20.00, context: 1000000 };
+const FABLE = { input: 10.00, output: 50.00, context: 1000000 };
+const SONNET_4X = { input: 3.00, output: 15.00, context: 1000000 };
 const SONNET = { input: 2.00, output: 10.00, context: 1000000 };
 const HAIKU = { input: 1.00, output: 5.00, context: 200000 };
 const GROK_46 = {
@@ -14,13 +17,25 @@ const GEMINI_PRO = {
     longContextThreshold: 200000, longContextInput: 4.00, longContextOutput: 18.00,
 };
 const GEMINI_FLASH = { input: 0.30, output: 2.50, context: 1000000 };
+// Gemini 3.6–3.8 Flash: launch price through 2026-12-31; Google lists $1.50/$7.50
+// from 2027-01-01. Update this entry before then.
+const GEMINI_FLASH_3X = { input: 0.75, output: 3.75, context: 1000000 };
+const GEMINI_35_FLASH = { input: 1.50, output: 9.00, context: 1000000 };
+const GEMINI_31_FLASH_LITE = { input: 0.25, output: 1.50, context: 1000000 };
+const GROK_BUILD = {
+    input: 1.00, output: 2.00, context: 256000,
+    longContextThreshold: 200000, longContextInput: 2.00, longContextOutput: 4.00,
+};
+const GPT_6_ASTRA = { input: 10.00, output: 50.00, context: 1050000 };
+const GPT_6_SOL = { input: 2.00, output: 10.00, context: 1050000 };
+const GPT_6_LUNA = { input: 0.10, output: 0.50, context: 1050000 };
 const GPT_SOL = { input: 4.00, output: 20.00, context: 1050000 };
 const GPT_TERRA = { input: 2.00, output: 12.00, context: 1050000 };
 const GPT_LUNA = { input: 0.20, output: 1.20, context: 1050000 };
 const GPT_54 = { input: 2.50, output: 15.00, context: 1000000 };
 const GPT_54_MINI = { input: 0.75, output: 4.50, context: 128000 };
 const GPT_54_NANO = { input: 0.20, output: 1.25, context: 128000 };
-// Pricing as of 2026-08-25 — USD per million tokens.
+// Pricing as of 2026-10-07 — USD per million tokens (checked against the four pages below).
 // First-party: platform.claude.com/docs/en/about-claude/pricing
 //              developers.openai.com/api/docs/models
 //              docs.x.ai/developers/models
@@ -28,23 +43,47 @@ const GPT_54_NANO = { input: 0.20, output: 1.25, context: 128000 };
 // Old keys stay as aliases so existing call sites don't throw.
 export const MODELS = {
     // Anthropic — live names + generic aliases (same rates)
+    'claude-fable-5.1': { ...FABLE },
+    'claude-fable-5': { ...FABLE },
+    'claude-mythos-5.1': { ...FABLE },
+    'claude-mythos-5': { ...FABLE },
+    'claude-opus-5.5': { ...OPUS_55 },
     'claude-opus-5': { ...OPUS },
+    'claude-opus-4.8': { ...OPUS },
     'claude-opus': { ...OPUS },
     'claude-opus-4.7': { ...OPUS },
+    'claude-opus-4.6': { ...OPUS },
+    'claude-opus-4.5': { ...OPUS },
+    'claude-sonnet-5.5': { ...SONNET },
     'claude-sonnet-5': { ...SONNET },
     'claude-sonnet': { ...SONNET },
+    'claude-sonnet-4.6': { ...SONNET_4X },
+    'claude-sonnet-4.5': { ...SONNET_4X },
     'claude-haiku-4.5': { ...HAIKU },
     'claude-haiku': { ...HAIKU },
     // xAI — flagship 4.6, cheap same-provider 4.3. 4.20 / fast are aliases.
+    'grok-4.7': { ...GROK_46 },
     'grok-4.6': { ...GROK_46 },
+    'grok-4.5': { ...GROK_46 },
+    'grok-build-0.1': { ...GROK_BUILD },
     'grok-4.3': { ...GROK_43 },
     'grok-4.20': { ...GROK_43 },
     'grok-4-1-fast': { ...GROK_43 },
     // Google
     'gemini-3.1-pro': { ...GEMINI_PRO },
+    'gemini-3.1-pro-preview': { ...GEMINI_PRO },
+    'gemini-3.8-flash': { ...GEMINI_FLASH_3X },
+    'gemini-3.7-flash': { ...GEMINI_FLASH_3X },
+    'gemini-3.6-flash': { ...GEMINI_FLASH_3X },
+    'gemini-3.5-flash': { ...GEMINI_35_FLASH },
+    'gemini-3.1-flash-lite': { ...GEMINI_31_FLASH_LITE },
     'gemini-3.5-flash-lite': { ...GEMINI_FLASH },
     'gemini-2.5-flash': { ...GEMINI_FLASH },
     // OpenAI — live 5.6 ladder. 5.4 family kept as aliases (old prices).
+    'gpt-6-astra': { ...GPT_6_ASTRA },
+    'gpt-6.1-sol': { ...GPT_6_SOL },
+    'gpt-6-sol': { ...GPT_6_SOL },
+    'gpt-6-luna': { ...GPT_6_LUNA },
     'gpt-5.6-sol': { ...GPT_SOL },
     'gpt-5.6-terra': { ...GPT_TERRA },
     'gpt-5.6-luna': { ...GPT_LUNA },
@@ -52,8 +91,28 @@ export const MODELS = {
     'gpt-5.4-mini': { ...GPT_54_MINI },
     'gpt-5.4-nano': { ...GPT_54_NANO },
 };
+/**
+ * Real API IDs → table keys, strictly: lowercase, drop a trailing date stamp
+ * (`-20250514`), and write version numbers with dots (`claude-opus-4-7` →
+ * `claude-opus-4.7`). No family guessing: an unknown version stays unknown,
+ * so a new model never silently gets an older model's price.
+ */
+export function canonicalModel(name) {
+    if (MODELS[name])
+        return name;
+    let n = name.trim().toLowerCase();
+    if (MODELS[n])
+        return n;
+    n = n.replace(/-\d{8}$/, '');
+    if (MODELS[n])
+        return n;
+    const dotted = n.replace(/(\d)-(\d)(?=$|-)/g, '$1.$2');
+    if (MODELS[dotted])
+        return dotted;
+    return n;
+}
 export function getModel(name) {
-    return MODELS[name] || MODELS[name.toLowerCase()];
+    return MODELS[canonicalModel(name)];
 }
 export function effectiveRate(tokens, info) {
     if (info.longContextThreshold !== undefined && tokens > info.longContextThreshold) {

package/dist/preflight.js

@@ -1,6 +1,6 @@
 import { slash } from './slash.js';
-import { getModel, MODELS, effectiveRate } from './models.js';
-import { PROVIDER_MODELS, providerOf } from './providers.js';
+import { getModel, MODELS, effectiveRate, canonicalModel } from './models.js';
+import { PROVIDER_MODELS, providerOf, NOT_ROUTE_TARGETS } from './providers.js';
 import { shouldRoute, isModelAllowed } from './config.js';
 /**
  * Compute cost of a prompt of `tokens` tokens on the given ModelInfo.
@@ -99,9 +99,12 @@ export function preflightRoute(content, model) {
         return null;
     const originalCost = computeCost(tokens, info);
     let cheapest = null;
+    const self = canonicalModel(model);
     for (const m of providerModels) {
-        if (m === model)
+        if (m === self)
             continue;
+        if (NOT_ROUTE_TARGETS.has(m))
+            continue; // specialised: never a target
         if (!isModelAllowed(m))
             continue; // user excluded this model
         const altInfo = getModel(m);
@@ -112,7 +115,10 @@ export function preflightRoute(content, model) {
         if (altInfo.input >= info.input)
             continue; // not cheaper
         const alt = buildAlternative(m, originalCost, tokens, altInfo);
-        if (!cheapest || alt.cost < cheapest.cost) {
+        // Tiny prompts round costs to the same value; break ties on the list price
+        // so the genuinely cheaper model wins (gpt-5.6-sol → gpt-6-luna, not 5.4-nano).
+        if (!cheapest || alt.cost < cheapest.cost ||
+            (alt.cost === cheapest.cost && altInfo.input < (getModel(cheapest.model)?.input ?? Infinity))) {
             cheapest = alt;
         }
     }

package/dist/providers.d.ts

@@ -8,4 +8,17 @@
  * strictly-cheaper hit wins (findCheapestRoute / preflightRoute).
  */
 export declare const PROVIDER_MODELS: Record<string, string[]>;
+/**
+ * Priced and grouped, but never a routing TARGET: specialised models a general
+ * call shouldn't be moved to (grok-build-0.1 is a coding-agent model).
+ */
+export declare const NOT_ROUTE_TARGETS: ReadonlySet<string>;
+/**
+ * Models `slash-tokens/auto` may rewrite a live request TO. Frozen at the
+ * 1.6.5 set so a patch release never changes where production calls go:
+ * models added in 1.6.6 (GPT-6, Gemini 3.1 Flash-Lite / 3.6–3.8 Flash, Grok
+ * 4.7 / 4.5, Claude 5.5 / Fable …) are priced and shown by preflight() /
+ * preflightRoute(), but /auto only starts routing to them in 1.7.0.
+ */
+export declare const AUTO_ROUTE_TARGETS: ReadonlySet<string>;
 export declare function providerOf(model: string): string | null;

package/dist/providers.js

@@ -1,3 +1,4 @@
+import { canonicalModel } from './models.js';
 /**
  * Provider groups — single source of truth.
  *
@@ -9,21 +10,49 @@
  */
 export const PROVIDER_MODELS = {
     Anthropic: [
-        'claude-opus-5', 'claude-opus', 'claude-opus-4.7',
-        'claude-sonnet-5', 'claude-sonnet',
+        'claude-fable-5.1', 'claude-fable-5', 'claude-mythos-5.1', 'claude-mythos-5',
+        'claude-opus-5', 'claude-opus', 'claude-opus-4.8', 'claude-opus-4.7',
+        'claude-opus-4.6', 'claude-opus-4.5', 'claude-opus-5.5',
+        'claude-sonnet-4.6', 'claude-sonnet-4.5',
+        'claude-sonnet-5.5', 'claude-sonnet-5', 'claude-sonnet',
         'claude-haiku', 'claude-haiku-4.5',
     ],
     OpenAI: [
+        'gpt-6-astra',
         'gpt-5.6-sol', 'gpt-5.6-terra',
-        'gpt-5.4', 'gpt-5.4-mini', 'gpt-5.4-nano',
-        'gpt-5.6-luna',
+        'gpt-5.4', 'gpt-6.1-sol', 'gpt-6-sol', 'gpt-5.4-mini', 'gpt-5.4-nano',
+        'gpt-5.6-luna', 'gpt-6-luna',
+    ],
+    xAI: ['grok-4.7', 'grok-4.6', 'grok-4.5', 'grok-4.3', 'grok-4.20', 'grok-4-1-fast', 'grok-build-0.1'],
+    Google: [
+        'gemini-3.1-pro', 'gemini-3.1-pro-preview',
+        'gemini-3.5-flash', 'gemini-3.8-flash', 'gemini-3.7-flash', 'gemini-3.6-flash',
+        'gemini-3.5-flash-lite', 'gemini-2.5-flash', 'gemini-3.1-flash-lite',
     ],
-    xAI: ['grok-4.6', 'grok-4.3', 'grok-4.20', 'grok-4-1-fast'],
-    Google: ['gemini-3.1-pro', 'gemini-3.5-flash-lite', 'gemini-2.5-flash'],
 };
+/**
+ * Priced and grouped, but never a routing TARGET: specialised models a general
+ * call shouldn't be moved to (grok-build-0.1 is a coding-agent model).
+ */
+export const NOT_ROUTE_TARGETS = new Set(['grok-build-0.1']);
+/**
+ * Models `slash-tokens/auto` may rewrite a live request TO. Frozen at the
+ * 1.6.5 set so a patch release never changes where production calls go:
+ * models added in 1.6.6 (GPT-6, Gemini 3.1 Flash-Lite / 3.6–3.8 Flash, Grok
+ * 4.7 / 4.5, Claude 5.5 / Fable …) are priced and shown by preflight() /
+ * preflightRoute(), but /auto only starts routing to them in 1.7.0.
+ */
+export const AUTO_ROUTE_TARGETS = new Set([
+    'claude-opus-5', 'claude-opus', 'claude-opus-4.7', 'claude-sonnet-5', 'claude-sonnet',
+    'claude-haiku', 'claude-haiku-4.5',
+    'gpt-5.6-sol', 'gpt-5.6-terra', 'gpt-5.4', 'gpt-5.4-mini', 'gpt-5.4-nano', 'gpt-5.6-luna',
+    'grok-4.6', 'grok-4.3', 'grok-4.20', 'grok-4-1-fast',
+    'gemini-3.1-pro', 'gemini-3.5-flash-lite', 'gemini-2.5-flash',
+]);
 export function providerOf(model) {
+    const id = canonicalModel(model);
     for (const [provider, models] of Object.entries(PROVIDER_MODELS)) {
-        if (models.includes(model))
+        if (models.includes(id))
             return provider;
     }
     return null;

package/dist/report.js

@@ -24,6 +24,7 @@ export function printReport(sites, filesScanned, timeMs, cwd) {
     if (sites.length === 0) {
         console.log(`${WHITE}  No AI API call sites detected.${R}`);
         console.log(`${GRAY}  Supported: OpenAI, Anthropic, Vercel AI, LangChain, Gemini, Bedrock, Grok${R}`);
+        console.log(`${GRAY}  Run this in a project that already calls an LLM.${R}`);
         console.log('');
         return;
     }

package/dist/slash.js

@@ -1,4 +1,5 @@
 import { getInstance, writeToMemory, ensureCapacity } from './wasm.js';
+import { canonicalModel } from './models.js';
 const WASM_INPUT_OFFSET = 4096;
 /**
  * Per-model calibration factors.
@@ -94,7 +95,18 @@ const WASM_INPUT_OFFSET = 4096;
  * Slash must NEVER under-report. Over-reporting is safe (go/no-go only).
  */
 const CALIBRATION = {
+    'claude-fable-5.1': 2.05,
+    'claude-fable-5': 2.05,
+    'claude-mythos-5.1': 2.05,
+    'claude-mythos-5': 2.05,
+    'claude-opus-5.5': 2.05,
     'claude-opus-5': 2.05,
+    'claude-opus-4.8': 2.05,
+    'claude-opus-4.6': 2.05,
+    'claude-opus-4.5': 2.05,
+    'claude-sonnet-5.5': 2.05,
+    'claude-sonnet-4.6': 2.05,
+    'claude-sonnet-4.5': 2.05,
     'claude-opus': 2.05,
     'claude-opus-4.7': 2.05,
     'claude-sonnet-5': 2.05,
@@ -102,9 +114,18 @@ const CALIBRATION = {
     'claude-haiku-4.5': 1.45,
     'claude-haiku': 1.45,
     'gemini-3.1-pro': 1.45,
+    'gemini-3.1-pro-preview': 1.45,
+    'gemini-3.8-flash': 1.45,
+    'gemini-3.7-flash': 1.45,
+    'gemini-3.6-flash': 1.45,
+    'gemini-3.5-flash': 1.45,
+    'gemini-3.1-flash-lite': 1.45,
     'gemini-3.5-flash-lite': 1.45,
     'gemini-2.5-flash': 1.45,
+    'grok-4.7': 1.15,
     'grok-4.6': 1.15,
+    'grok-4.5': 1.15,
+    'grok-build-0.1': 1.15,
     'grok-4.3': 1.15,
     'grok-4.20': 1.15,
     'grok-4-1-fast': 1.15,
@@ -141,7 +162,10 @@ export function slash(content, model) {
     const raw = instance.exports.estimate_tokens(WASM_INPUT_OFFSET, len);
     if (!model)
         return raw;
-    const factor = CALIBRATION[model] ?? DEFAULT_UNKNOWN_MODEL_FACTOR;
+    // GPT-6 (gpt-6-astra / 6.1-sol / 6-sol / 6-luna) is deliberately NOT in
+    // CALIBRATION: a new generation with an unbenchmarked tokenizer takes the
+    // conservative default (never under-report) until a bench run adds it.
+    const factor = CALIBRATION[canonicalModel(model)] ?? DEFAULT_UNKNOWN_MODEL_FACTOR;
     return factor === 1.0 ? raw : Math.ceil(raw * factor);
 }
 /**

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "slash-tokens",
-  "version": "1.6.5",
+  "version": "1.6.6",
   "description": "Token Optimization for Context Engineers. 4.8 KB WASM. Sub-millisecond. Zero dependencies.",
   "main": "dist/index.js",
   "module": "dist/index.js",
@@ -26,8 +26,10 @@
   ],
   "sideEffects": false,
   "scripts": {
+    "prepublishOnly": "npm run build",
     "build": "npx tsc && bun build src/cli.ts --outfile=dist/cli.js --target=node",
-    "test": "bun test"
+    "test": "bun test",
+    "check:engines": "node scripts/check-engines.mjs"
   },
   "author": "wolfejam",
   "license": "MIT",
@@ -54,5 +56,8 @@
     "@types/node": "^25.5.0",
     "js-tiktoken": "^1.0.21",
     "typescript": "^6.0.2"
+  },
+  "engines": {
+    "node": ">=22.0.0"
   }
 }