vibezcheck 0.4.2 → 0.5.3

package/dist/meter/index.mjs

@@ -259,12 +259,12 @@ function extractOpenAIResponseUsage(response) {
 function inspectOpenAIStreamChunk(chunk) {
   if (!chunk || typeof chunk !== "object") return {};
   const model = chunk.model;
-  if (chunk.usage) {
-    const rawUsage = chunk.usage;
-    const inputTokens = rawUsage.prompt_tokens ?? rawUsage.input_tokens ?? 0;
-    const outputTokens = rawUsage.completion_tokens ?? rawUsage.output_tokens ?? 0;
-    const reasoningTokens = rawUsage.completion_tokens_details?.reasoning_tokens ?? rawUsage.output_token_details?.reasoning_tokens ?? 0;
-    const cachedTokens = rawUsage.prompt_tokens_details?.cached_tokens ?? rawUsage.input_token_details?.cached_tokens ?? 0;
+  const rawUsage = chunk.usage || chunk.token_usage || chunk.x_groq?.usage || chunk.usageMetadata;
+  if (rawUsage) {
+    const inputTokens = rawUsage.prompt_tokens ?? rawUsage.input_tokens ?? rawUsage.promptTokens ?? rawUsage.inputTokens ?? rawUsage.promptTokenCount ?? 0;
+    const outputTokens = rawUsage.completion_tokens ?? rawUsage.output_tokens ?? rawUsage.completionTokens ?? rawUsage.outputTokens ?? rawUsage.candidatesTokenCount ?? 0;
+    const reasoningTokens = rawUsage.completion_tokens_details?.reasoning_tokens ?? rawUsage.output_token_details?.reasoning_tokens ?? rawUsage.reasoning_tokens ?? rawUsage.reasoningTokens ?? rawUsage.thoughtsTokenCount ?? 0;
+    const cachedTokens = rawUsage.prompt_tokens_details?.cached_tokens ?? rawUsage.input_token_details?.cached_tokens ?? rawUsage.prompt_cache_hit_tokens ?? rawUsage.cached_tokens ?? rawUsage.cachedTokens ?? rawUsage.cachedContentTokenCount ?? 0;
     return {
       model,
       usage: {
@@ -279,11 +279,11 @@ function inspectOpenAIStreamChunk(chunk) {
   }
   if (chunk.type === "response.completed" || chunk.type === "response.done") {
     if (chunk.response?.usage) {
-      const rawUsage = chunk.response.usage;
-      const inputTokens = rawUsage.input_tokens ?? 0;
-      const outputTokens = rawUsage.output_tokens ?? 0;
-      const reasoningTokens = rawUsage.output_token_details?.reasoning_tokens ?? 0;
-      const cachedTokens = rawUsage.input_token_details?.cached_tokens ?? 0;
+      const rawUsage2 = chunk.response.usage;
+      const inputTokens = rawUsage2.input_tokens ?? 0;
+      const outputTokens = rawUsage2.output_tokens ?? 0;
+      const reasoningTokens = rawUsage2.output_token_details?.reasoning_tokens ?? 0;
+      const cachedTokens = rawUsage2.input_token_details?.cached_tokens ?? 0;
       return {
         model: chunk.response.model || model,
         usage: {
@@ -448,7 +448,7 @@ function detectAndExtractUsage(response, fallbackModel, fallbackProvider) {
 
 // src/pricing/table.ts
 var MODEL_PRICING_TABLE = {
-  // --- OpenAI ---
+  // --- OpenAI (Modern & Reasoning) ---
   "gpt-5.6-sol": { inputPer1M: 4, outputPer1M: 20, cachedInputPer1M: 0.4 },
   "gpt-5.6-terra": { inputPer1M: 2, outputPer1M: 12, cachedInputPer1M: 0.2 },
   "gpt-5.6-luna": { inputPer1M: 0.2, outputPer1M: 1.2, cachedInputPer1M: 0.02 },
@@ -460,11 +460,22 @@ var MODEL_PRICING_TABLE = {
   "o3-mini": { inputPer1M: 1.1, outputPer1M: 4.4, cachedInputPer1M: 0.55 },
   "gpt-4o": { inputPer1M: 2.5, outputPer1M: 10, cachedInputPer1M: 1.25 },
   "gpt-4o-mini": { inputPer1M: 0.15, outputPer1M: 0.6, cachedInputPer1M: 0.075 },
+  "gpt-4.5": { inputPer1M: 75, outputPer1M: 150, cachedInputPer1M: 37.5 },
+  "gpt-4.5-preview": { inputPer1M: 75, outputPer1M: 150, cachedInputPer1M: 37.5 },
+  "chatgpt-4o-latest": { inputPer1M: 5, outputPer1M: 15 },
   "gpt-4.1": { inputPer1M: 2, outputPer1M: 8, cachedInputPer1M: 1 },
   "gpt-4.1-nano": { inputPer1M: 0.1, outputPer1M: 0.4, cachedInputPer1M: 0.05 },
+  // --- OpenAI (Legacy & Backward Compatibility) ---
+  "gpt-4-turbo": { inputPer1M: 10, outputPer1M: 30, cachedInputPer1M: 5 },
+  "gpt-4-turbo-preview": { inputPer1M: 10, outputPer1M: 30, cachedInputPer1M: 5 },
+  "gpt-4": { inputPer1M: 30, outputPer1M: 60 },
+  "gpt-4-32k": { inputPer1M: 60, outputPer1M: 120 },
+  "gpt-3.5-turbo": { inputPer1M: 0.5, outputPer1M: 1.5 },
+  "gpt-3.5-turbo-16k": { inputPer1M: 3, outputPer1M: 4 },
   "text-embedding-3-small": { inputPer1M: 0.02, outputPer1M: 0 },
   "text-embedding-3-large": { inputPer1M: 0.13, outputPer1M: 0 },
-  // --- Anthropic ---
+  "text-embedding-ada-002": { inputPer1M: 0.1, outputPer1M: 0 },
+  // --- Anthropic (Modern & Extended Thinking) ---
   "claude-3-7-sonnet": { inputPer1M: 0.59, outputPer1M: 2.93, cachedInputPer1M: 0.3 },
   "claude-sonnet-5": { inputPer1M: 2, outputPer1M: 10, cachedInputPer1M: 0.3 },
   "claude-3-5-sonnet": { inputPer1M: 3, outputPer1M: 15, cachedInputPer1M: 0.3 },
@@ -472,48 +483,145 @@ var MODEL_PRICING_TABLE = {
   "haiku-4.5": { inputPer1M: 1, outputPer1M: 5, cachedInputPer1M: 0.1 },
   "claude-opus-5": { inputPer1M: 5, outputPer1M: 25, cachedInputPer1M: 1.5 },
   "claude-3-opus": { inputPer1M: 15, outputPer1M: 75, cachedInputPer1M: 1.5 },
-  // --- Google Gemini ---
+  // --- Anthropic (Legacy & Backward Compatibility) ---
+  "claude-3-sonnet": { inputPer1M: 3, outputPer1M: 15, cachedInputPer1M: 0.3 },
+  "claude-3-haiku": { inputPer1M: 0.25, outputPer1M: 1.25, cachedInputPer1M: 0.025 },
+  "claude-2.1": { inputPer1M: 8, outputPer1M: 24 },
+  "claude-2.0": { inputPer1M: 8, outputPer1M: 24 },
+  "claude-instant-1.2": { inputPer1M: 1.63, outputPer1M: 5.51 },
+  // --- Google Gemini (Modern & Thoughts) ---
   "gemini-3.7-flash": { inputPer1M: 0.75, outputPer1M: 3.75, cachedInputPer1M: 0.18 },
   "gemini-3.1-pro": { inputPer1M: 2, outputPer1M: 12, cachedInputPer1M: 0.5 },
   "gemini-3.5-flash": { inputPer1M: 1.5, outputPer1M: 9, cachedInputPer1M: 0.38 },
   "gemini-3.1-flash-lite": { inputPer1M: 0.25, outputPer1M: 1.5, cachedInputPer1M: 0.06 },
+  "gemini-2.5-pro": { inputPer1M: 1.25, outputPer1M: 5, cachedInputPer1M: 0.3125 },
+  "gemini-2.5-flash": { inputPer1M: 0.15, outputPer1M: 0.6, cachedInputPer1M: 0.0375 },
   "gemini-2.0-flash": { inputPer1M: 0.1, outputPer1M: 0.4, cachedInputPer1M: 0.025 },
+  "gemini-2.0-flash-lite": { inputPer1M: 0.075, outputPer1M: 0.3, cachedInputPer1M: 0.01875 },
+  "gemini-2.0-pro": { inputPer1M: 1.25, outputPer1M: 5, cachedInputPer1M: 0.3125 },
+  // --- Google Gemini (Legacy & Backward Compatibility) ---
   "gemini-1.5-pro": { inputPer1M: 1.25, outputPer1M: 5, cachedInputPer1M: 0.3125 },
   "gemini-1.5-flash": { inputPer1M: 0.075, outputPer1M: 0.3, cachedInputPer1M: 0.01875 },
+  "gemini-1.5-flash-8b": { inputPer1M: 0.0375, outputPer1M: 0.15, cachedInputPer1M: 9375e-6 },
+  "gemini-1.0-pro": { inputPer1M: 0.5, outputPer1M: 1.5 },
   // --- xAI Grok ---
   "grok-4.6": { inputPer1M: 3, outputPer1M: 15 },
+  "grok-3": { inputPer1M: 3, outputPer1M: 15 },
+  "grok-3-mini": { inputPer1M: 0.3, outputPer1M: 1.5 },
   "grok-2": { inputPer1M: 2, outputPer1M: 10 },
   "grok-2-vision": { inputPer1M: 2, outputPer1M: 10 },
   "grok-beta": { inputPer1M: 5, outputPer1M: 15 },
   // --- Mistral ---
   "mistral-large-3": { inputPer1M: 2, outputPer1M: 6 },
   "mistral-large-latest": { inputPer1M: 2, outputPer1M: 6 },
+  "mistral-large-2411": { inputPer1M: 2, outputPer1M: 6 },
   "codestral-latest": { inputPer1M: 0.3, outputPer1M: 0.9 },
+  "mistral-medium-latest": { inputPer1M: 2.7, outputPer1M: 8.1 },
   "mistral-small-latest": { inputPer1M: 0.2, outputPer1M: 0.6 },
   "ministral-8b-latest": { inputPer1M: 0.1, outputPer1M: 0.1 },
-  // --- Groq LPUs ---
+  "ministral-3b-latest": { inputPer1M: 0.04, outputPer1M: 0.04 },
+  "open-mistral-7b": { inputPer1M: 0.2, outputPer1M: 0.2 },
+  "open-mixtral-8x7b": { inputPer1M: 0.7, outputPer1M: 0.7 },
+  "open-mixtral-8x22b": { inputPer1M: 2, outputPer1M: 6 },
+  // --- Groq & Meta Llama LPUs ---
   "llama-3.3-70b-versatile": { inputPer1M: 0.59, outputPer1M: 0.79 },
+  "llama-3.1-405b": { inputPer1M: 3, outputPer1M: 3 },
+  "llama-3.1-70b-versatile": { inputPer1M: 0.59, outputPer1M: 0.79 },
   "llama-3.1-8b-instant": { inputPer1M: 0.05, outputPer1M: 0.08 },
+  "llama-3.2-1b-preview": { inputPer1M: 0.04, outputPer1M: 0.04 },
+  "llama-3.2-3b-preview": { inputPer1M: 0.06, outputPer1M: 0.06 },
+  "llama-3.2-11b-vision": { inputPer1M: 0.18, outputPer1M: 0.18 },
+  "llama-3.2-90b-vision": { inputPer1M: 0.9, outputPer1M: 0.9 },
   "deepseek-r1-distill-llama-70b": { inputPer1M: 0.75, outputPer1M: 0.99 },
+  "deepseek-r1-distill-qwen-32b": { inputPer1M: 0.49, outputPer1M: 0.49 },
   "qwen-2.5-32b": { inputPer1M: 0.29, outputPer1M: 0.39 },
+  "qwen-2.5-72b": { inputPer1M: 0.35, outputPer1M: 0.4 },
+  "mixtral-8x7b-32768": { inputPer1M: 0.24, outputPer1M: 0.24 },
+  "gemma2-9b-it": { inputPer1M: 0.2, outputPer1M: 0.2 },
   // --- DeepSeek ---
   "deepseek-v4-pro": { inputPer1M: 0.66, outputPer1M: 1.98, cachedInputPer1M: 0.15 },
   "deepseek-v4-flash": { inputPer1M: 0.22, outputPer1M: 0.66, cachedInputPer1M: 0.05 },
-  "deepseek-chat": { inputPer1M: 0.22, outputPer1M: 0.66, cachedInputPer1M: 0.05 },
-  "deepseek-reasoner": { inputPer1M: 0.66, outputPer1M: 1.98, cachedInputPer1M: 0.15 },
+  "deepseek-v3": { inputPer1M: 0.14, outputPer1M: 0.28, cachedInputPer1M: 0.014 },
+  "deepseek-chat": { inputPer1M: 0.14, outputPer1M: 0.28, cachedInputPer1M: 0.014 },
+  "deepseek-r1": { inputPer1M: 0.55, outputPer1M: 2.19, cachedInputPer1M: 0.14 },
+  "deepseek-reasoner": { inputPer1M: 0.55, outputPer1M: 2.19, cachedInputPer1M: 0.14 },
   // --- Cohere ---
   "command-r-plus": { inputPer1M: 2.5, outputPer1M: 10 },
-  "command-r": { inputPer1M: 0.15, outputPer1M: 0.6 }
+  "command-r": { inputPer1M: 0.15, outputPer1M: 0.6 },
+  "command": { inputPer1M: 1, outputPer1M: 2 },
+  "command-light": { inputPer1M: 0.3, outputPer1M: 0.6 },
+  "embed-english-v3.0": { inputPer1M: 0.1, outputPer1M: 0 },
+  // --- Perplexity ---
+  "sonar": { inputPer1M: 1, outputPer1M: 1 },
+  "sonar-pro": { inputPer1M: 3, outputPer1M: 15 },
+  "sonar-reasoning": { inputPer1M: 1, outputPer1M: 5 },
+  "sonar-reasoning-pro": { inputPer1M: 2, outputPer1M: 8 }
+};
+var MODEL_ALIASES = {
+  // OpenAI Shorthands & Aliases
+  "gpt4": "gpt-4",
+  "gpt4o": "gpt-4o",
+  "gpt-4-preview": "gpt-4-turbo",
+  "gpt-4-0125-preview": "gpt-4-turbo",
+  "gpt-4-1106-preview": "gpt-4-turbo",
+  "gpt-3.5": "gpt-3.5-turbo",
+  "gpt-3.5-turbo-0125": "gpt-3.5-turbo",
+  "gpt-3.5-turbo-1106": "gpt-3.5-turbo",
+  "gpt-3.5-turbo-16k-0613": "gpt-3.5-turbo-16k",
+  // Anthropic Shorthands
+  "sonnet": "claude-3-5-sonnet",
+  "sonnet-3.7": "claude-3-7-sonnet",
+  "sonnet-3.5": "claude-3-5-sonnet",
+  "haiku": "claude-3-5-haiku",
+  "haiku-3.5": "claude-3-5-haiku",
+  "opus": "claude-3-opus",
+  "claude-2": "claude-2.0",
+  "claude-instant": "claude-instant-1.2",
+  // DeepSeek Shorthands
+  "r1": "deepseek-r1",
+  "v3": "deepseek-v3",
+  // Gemini Shorthands
+  "flash": "gemini-2.0-flash",
+  "pro": "gemini-1.5-pro",
+  // Meta / Groq Shorthands
+  "llama-3.3-70b": "llama-3.3-70b-versatile",
+  "llama-3.1-70b": "llama-3.1-70b-versatile",
+  "llama-3.1-8b": "llama-3.1-8b-instant",
+  "llama-3-70b": "llama-3.1-70b-versatile",
+  "llama-3-8b": "llama-3.1-8b-instant",
+  // Mistral Shorthands
+  "codestral": "codestral-latest",
+  "mistral-large": "mistral-large-latest",
+  "mistral-small": "mistral-small-latest"
 };
 var customPricingRegistry = {};
 function normalizeModelKey(rawModel) {
   if (!rawModel) return "unknown";
   let model = rawModel.toLowerCase().trim();
+  model = model.replace(/^[a-z0-9_-]+\.(anthropic|meta|amazon|cohere|mistral|ai21)\./i, "");
+  model = model.replace(/^(anthropic|meta|amazon|cohere|mistral|ai21)\./i, "");
+  model = model.replace(/-v\d+(:\d+)?$/, "");
+  model = model.replace(/:\d+$/, "");
   if (model.includes("/")) {
-    model = model.split("/")[1] || model;
+    model = model.split("/").slice(1).join("/");
   }
+  model = model.replace(/:(latest|free|beta)$/, "");
   model = model.replace(/-\d{8}$/, "");
   model = model.replace(/-\d{4}-\d{2}-\d{2}$/, "");
+  model = model.replace(/llama(\d+)-(\d+)-/g, "llama-$1.$2-");
+  model = model.replace(/llama(\d+)\.(\d+)-/g, "llama-$1.$2-");
+  if (MODEL_ALIASES[model]) {
+    return MODEL_ALIASES[model];
+  }
+  if (!MODEL_PRICING_TABLE[model]) {
+    const withoutInstruct = model.replace(/-(instruct|chat|preview)$/, "");
+    if (MODEL_PRICING_TABLE[withoutInstruct]) {
+      return withoutInstruct;
+    }
+    if (MODEL_ALIASES[withoutInstruct]) {
+      return MODEL_ALIASES[withoutInstruct];
+    }
+  }
   return model;
 }
 function getModelPricing(modelName) {
@@ -530,6 +638,10 @@ function getModelPricing(modelName) {
   if (MODEL_PRICING_TABLE[modelName]) {
     return MODEL_PRICING_TABLE[modelName];
   }
+  const alias = MODEL_ALIASES[modelName.toLowerCase().trim()];
+  if (alias && MODEL_PRICING_TABLE[alias]) {
+    return MODEL_PRICING_TABLE[alias];
+  }
   return {
     inputPer1M: 1,
     outputPer1M: 3,
@@ -552,35 +664,57 @@ function calculateCost(params) {
   } else {
     rates = getModelPricing(params.model);
   }
-  const inputTokens = params.inputTokens ?? 0;
-  const outputTokens = params.outputTokens ?? 0;
-  const reasoningTokens = params.reasoningTokens ?? 0;
-  const cachedTokens = params.cachedTokens ?? 0;
-  const regularInputTokens = Math.max(0, inputTokens - cachedTokens);
-  const regularInputCost = regularInputTokens / 1e6 * rates.inputPer1M;
-  const cachedRate = rates.cachedInputPer1M ?? rates.inputPer1M * 0.15;
-  const cachedInputCost = cachedTokens / 1e6 * cachedRate;
-  const inputCostUSD = regularInputCost + cachedInputCost;
-  const outputCostUSD = outputTokens / 1e6 * rates.outputPer1M;
-  const reasoningRate = rates.reasoningPer1M ?? rates.outputPer1M;
-  const reasoningCostUSD = reasoningTokens / 1e6 * reasoningRate;
-  const standardCacheCost = cachedTokens / 1e6 * rates.inputPer1M;
-  const cachedDiscountUSD = Math.max(0, standardCacheCost - cachedInputCost);
-  let totalUSD = inputCostUSD + outputCostUSD;
+  const inputTokens = BigInt(Math.max(0, params.inputTokens ?? 0));
+  const outputTokens = BigInt(Math.max(0, params.outputTokens ?? 0));
+  const reasoningTokens = BigInt(Math.max(0, params.reasoningTokens ?? 0));
+  const cachedTokens = BigInt(Math.max(0, params.cachedTokens ?? 0));
+  const regularInputTokens = inputTokens > cachedTokens ? inputTokens - cachedTokens : 0n;
+  const rateInputNano = BigInt(Math.round(rates.inputPer1M * 1e3));
+  const cachedPer1M = rates.cachedInputPer1M ?? rates.inputPer1M * 0.15;
+  const rateCachedNano = BigInt(Math.round(cachedPer1M * 1e3));
+  const rateOutputNano = BigInt(Math.round(rates.outputPer1M * 1e3));
+  const reasoningPer1M = rates.reasoningPer1M ?? rates.outputPer1M;
+  const rateReasoningNano = BigInt(Math.round(reasoningPer1M * 1e3));
+  const regularInputCostNano = regularInputTokens * rateInputNano;
+  const cachedInputCostNano = cachedTokens * rateCachedNano;
+  const inputCostNano = regularInputCostNano + cachedInputCostNano;
+  const outputCostNano = outputTokens * rateOutputNano;
+  const reasoningCostNano = reasoningTokens * rateReasoningNano;
+  const standardCacheCostNano = cachedTokens * rateInputNano;
+  const cachedDiscountNano = standardCacheCostNano > cachedInputCostNano ? standardCacheCostNano - cachedInputCostNano : 0n;
+  const wholesaleNano = inputCostNano + outputCostNano;
   const markup = params.markupMultiplier ?? 1;
-  let retailUSD = void 0;
+  let billedNano = wholesaleNano;
+  let hasRetail = false;
   if (markup !== 1 || params.minimumChargeUSD !== void 0) {
-    retailUSD = totalUSD * markup;
-    if (params.minimumChargeUSD !== void 0 && retailUSD < params.minimumChargeUSD) {
-      retailUSD = params.minimumChargeUSD;
+    hasRetail = true;
+    const markupMultiplierNano = BigInt(Math.round(markup * 1e3));
+    billedNano = wholesaleNano * markupMultiplierNano / 1000n;
+    if (params.minimumChargeUSD !== void 0 && params.minimumChargeUSD > 0) {
+      const minChargeNano = BigInt(Math.round(params.minimumChargeUSD * 1e9));
+      if (billedNano < minChargeNano) {
+        billedNano = minChargeNano;
+      }
     }
   }
+  const inputCostUSD = Number(inputCostNano) / 1e9;
+  const outputCostUSD = Number(outputCostNano) / 1e9;
+  const reasoningCostUSD = Number(reasoningCostNano) / 1e9;
+  const cachedDiscountUSD = Number(cachedDiscountNano) / 1e9;
+  const totalUSD = Number(wholesaleNano) / 1e9;
+  const wholesaleTotalUSD = totalUSD;
+  const billedUSD = Number(billedNano) / 1e9;
+  const profitUSD = Math.max(0, billedUSD - wholesaleTotalUSD);
+  const retailUSD = hasRetail ? billedUSD : void 0;
   return {
     inputCostUSD: Number(inputCostUSD.toFixed(8)),
     outputCostUSD: Number(outputCostUSD.toFixed(8)),
-    reasoningCostUSD: reasoningTokens > 0 ? Number(reasoningCostUSD.toFixed(8)) : void 0,
-    cachedDiscountUSD: cachedTokens > 0 ? Number(cachedDiscountUSD.toFixed(8)) : void 0,
+    reasoningCostUSD: reasoningTokens > 0n ? Number(reasoningCostUSD.toFixed(8)) : void 0,
+    cachedDiscountUSD: cachedTokens > 0n ? Number(cachedDiscountUSD.toFixed(8)) : void 0,
     totalUSD: Number(totalUSD.toFixed(8)),
+    wholesaleTotalUSD: Number(wholesaleTotalUSD.toFixed(8)),
+    billedUSD: Number(billedUSD.toFixed(8)),
+    profitUSD: Number(profitUSD.toFixed(8)),
     retailUSD: retailUSD !== void 0 ? Number(retailUSD.toFixed(8)) : void 0,
     currency: rates.currency || "USD"
   };
@@ -786,7 +920,7 @@ function wrapUniversalStream(stream, options = {}, onComplete) {
     return wrapGeminiStream(stream, options, onComplete);
   }
   if (Symbol.asyncIterator in stream) {
-    if (options.provider === "anthropic") {
+    if (options.provider === "anthropic" || options.model?.toLowerCase().includes("claude")) {
       return wrapAnthropicStream(stream, options, onComplete);
     }
     return wrapOpenAIStream(stream, options, onComplete);
@@ -808,7 +942,7 @@ var VibezMeter = class {
       this.stripeClient = new Stripe(key, {
         appInfo: {
           name: "vibezcheck",
-          version: "0.4.2",
+          version: "0.5.3",
           url: "https://vibezcheck.xyz"
         }
       });