Improve multi-editor model-aware AI cost tracking and aggregation.

This updates pricing/model detection across Cursor, VS Code, Windsurf, Antigravity, and Claude Code while adding a single aggregated AI summary output for budgeting and dashboard visibility.

Made-with: Cursor

Author: bhav66d

packages/core/src/cost.ts

@@ -1,22 +1,14 @@
 import { UsageEvent, CostBreakdown } from './types';
+import { calculateCost } from './pricing';
 
 export class CostCalculator {
-  private readonly modelCosts: Record<string, { input: number; output: number; cacheRead?: number; cacheWrite?: number }> = {
-    'gpt-4': { input: 0.03, output: 0.06 },
-    'gpt-4-turbo': { input: 0.01, output: 0.03 },
-    'gpt-3.5-turbo': { input: 0.0005, output: 0.0015 },
-    'claude-3-opus': { input: 0.015, output: 0.075 },
-    'claude-3-sonnet': { input: 0.003, output: 0.015 },
-    'claude-3-haiku': { input: 0.00025, output: 0.00125 }
-  };
-
   calculateCost(events: UsageEvent[]): CostBreakdown {
     const now = Date.now();
     const startMs = now - (24 * 60 * 60 * 1000);
     const endMs = now;
 
     const dailyEvents = events.filter(e => e.timestampMs >= startMs && e.timestampMs <= endMs);
-    
+
     const totalCents = dailyEvents.reduce((sum, event) => sum + event.cost.totalCents, 0);
     const modelBreakdown: Record<string, number> = {};
     const tokenBreakdown = {
@@ -50,23 +42,19 @@ export class CostCalculator {
     };
   }
 
+  /**
+   * Fill in missing cost data for a usage event using live pricing.
+   * Returns cost in cents.
+   */
   estimateMissingCost(event: UsageEvent): number {
     if (event.cost.totalCents > 0) {
       return event.cost.totalCents;
     }
 
-    const modelCost = this.modelCosts[event.model.toLowerCase()];
-    if (!modelCost) {
-      return 0;
-    }
-
-    const inputCost = (event.tokenUsage.inputTokens / 1000) * modelCost.input;
-    const outputCost = (event.tokenUsage.outputTokens / 1000) * modelCost.output;
-    const cacheReadCost = event.tokenUsage.cacheReadTokens ? 
-      (event.tokenUsage.cacheReadTokens / 1000) * (modelCost.cacheRead || modelCost.input * 0.1) : 0;
-    const cacheWriteCost = event.tokenUsage.cacheWriteTokens ? 
-      (event.tokenUsage.cacheWriteTokens / 1000) * (modelCost.cacheWrite || modelCost.input * 0.5) : 0;
-
-    return Math.round((inputCost + outputCost + cacheReadCost + cacheWriteCost) * 100);
+    return calculateCost(
+      event.tokenUsage.inputTokens,
+      event.tokenUsage.outputTokens,
+      event.model
+    );
   }
 }