feat(codex): compute Codex credit usage (#408, #495)

mac/Sources/CodeBurnMenubar/AppStore.swift

@@ -1142,7 +1142,7 @@ enum SubscriptionLoadState: Sendable, Equatable {
 }
 
 enum DisplayMetric: String {
-    case cost, tokens, totalTokens, iconOnly
+    case cost, tokens, totalTokens, credits, iconOnly
 }
 
 enum InsightMode: String, CaseIterable, Identifiable {

mac/Sources/CodeBurnMenubar/CodeBurnApp.swift

@@ -735,6 +735,9 @@ final class AppDelegate: NSObject, NSApplicationDelegate, NSPopoverDelegate {
             } else if store.displayMetric == .totalTokens, let p = menubarPayload?.current {
                 let total = formatTokensMenubar(Double(p.inputTokens + p.outputTokens))
                 valueText = compact ? "\(total)\(suffix)" : " \(total) tok\(suffix)"
+            } else if store.displayMetric == .credits, let p = menubarPayload?.current {
+                let credits = formatTokensMenubar((p.codexCredits ?? 0).rounded())
+                valueText = compact ? "\(credits)cr\(suffix)" : " \(credits) credits\(suffix)"
             } else {
                 let fallback = compact ? "$-" : "$—"
                 let formatted = menubarPayload?.current.cost

mac/Sources/CodeBurnMenubar/Data/MenubarPayload.swift

@@ -113,6 +113,8 @@ struct CurrentBlock: Codable, Sendable {
     let inputTokens: Int
     let outputTokens: Int
     let cacheHitPercent: Double
+    /// Codex credits consumed in the period (nil on payloads from older builds).
+    let codexCredits: Double?
     let topActivities: [ActivityEntry]
     let topModels: [ModelEntry]
     let localModelSavings: LocalModelSavings
@@ -131,7 +133,7 @@ struct CurrentBlock: Codable, Sendable {
 extension CurrentBlock {
     enum CodingKeys: String, CodingKey {
         case label, cost, calls, sessions, oneShotRate, inputTokens, outputTokens,
-             cacheHitPercent, topActivities, topModels, localModelSavings, providers, topProjects,
+             cacheHitPercent, codexCredits, topActivities, topModels, localModelSavings, providers, topProjects,
              modelEfficiency, topSessions, retryTax, routingWaste,
              tools, skills, subagents, mcpServers
     }
@@ -145,6 +147,7 @@ extension CurrentBlock {
         inputTokens = try c.decode(Int.self, forKey: .inputTokens)
         outputTokens = try c.decode(Int.self, forKey: .outputTokens)
         cacheHitPercent = try c.decodeIfPresent(Double.self, forKey: .cacheHitPercent) ?? 0
+        codexCredits = try c.decodeIfPresent(Double.self, forKey: .codexCredits)
         topActivities = try c.decodeIfPresent([ActivityEntry].self, forKey: .topActivities) ?? []
         topModels = try c.decodeIfPresent([ModelEntry].self, forKey: .topModels) ?? []
         localModelSavings = try c.decodeIfPresent(LocalModelSavings.self, forKey: .localModelSavings) ?? LocalModelSavings(totalUSD: 0, calls: 0, byModel: [], byProvider: [])
@@ -368,6 +371,7 @@ extension MenubarPayload {
             inputTokens: 0,
             outputTokens: 0,
             cacheHitPercent: 0,
+            codexCredits: nil,
             topActivities: [],
             topModels: [],
             localModelSavings: LocalModelSavings(totalUSD: 0, calls: 0, byModel: [], byProvider: []),

mac/Sources/CodeBurnMenubar/Views/SettingsView.swift

@@ -91,6 +91,7 @@ private struct GeneralSettingsTab: View {
                     Text("Cost ($)").tag(DisplayMetric.cost)
                     Text("Tokens (↑↓)").tag(DisplayMetric.tokens)
                     Text("Total Tokens").tag(DisplayMetric.totalTokens)
+                    Text("Credits (Codex)").tag(DisplayMetric.credits)
                     Text("Icon Only").tag(DisplayMetric.iconOnly)
                 }
                 Picker("Period", selection: Binding(

src/codex-credits.ts

@@ -0,0 +1,57 @@
+// Codex credit pricing. ChatGPT/Codex subscription users consume *credits*, a
+// separate unit from API dollars: usage is billed as "credits per million
+// tokens" at per-model rates that differ from the API USD pricing CodeBurn uses
+// for cost. This module computes credit consumption from token counts so the
+// app can show usage in credits (issues #408 and #495).
+//
+// Rates are credits per 1,000,000 tokens, from
+// https://developers.openai.com/codex/pricing#credits-overview
+// (cached input is the cheaper rate applied to cache-read tokens).
+
+export type CodexCreditRate = {
+  input: number
+  cachedInput: number
+  output: number
+}
+
+const CREDITS_PER_MILLION: Record<string, CodexCreditRate> = {
+  'gpt-5.5': { input: 125, cachedInput: 12.5, output: 750 },
+  'gpt-5.4': { input: 62.5, cachedInput: 6.25, output: 375 },
+  'gpt-5.4-mini': { input: 18.75, cachedInput: 1.875, output: 113 },
+}
+
+/// Resolve the credit rate for a Codex model name, tolerating suffix variants
+/// (e.g. "gpt-5.5-codex"). Returns null when the model has no known credit rate.
+export function codexCreditRate(model: string): CodexCreditRate | null {
+  const m = model.toLowerCase()
+  if (m.includes('5.4') && m.includes('mini')) return CREDITS_PER_MILLION['gpt-5.4-mini']!
+  if (m.includes('5.4')) return CREDITS_PER_MILLION['gpt-5.4']!
+  if (m.includes('5.5')) return CREDITS_PER_MILLION['gpt-5.5']!
+  return null
+}
+
+export type CodexCreditTokens = {
+  /// Non-cached input tokens (CodeBurn normalizes Codex to Anthropic semantics,
+  /// so this excludes cache-read tokens).
+  inputTokens: number
+  /// Cache-read (cached input) tokens, billed at the cheaper cached rate.
+  cachedReadTokens: number
+  outputTokens: number
+  /// Reasoning tokens are billed as output, matching CodeBurn's cost model.
+  reasoningTokens?: number
+}
+
+/// Credits consumed for one Codex usage record. Returns null when the model has
+/// no known credit rate (caller decides how to surface "unknown").
+export function codexCredits(model: string, tokens: CodexCreditTokens): number | null {
+  const rate = codexCreditRate(model)
+  if (!rate) return null
+  const safe = (n: number) => (Number.isFinite(n) && n > 0 ? n : 0)
+  const PER_MILLION = 1_000_000
+  const output = safe(tokens.outputTokens) + safe(tokens.reasoningTokens ?? 0)
+  return (
+    (safe(tokens.inputTokens) / PER_MILLION) * rate.input +
+    (safe(tokens.cachedReadTokens) / PER_MILLION) * rate.cachedInput +
+    (output / PER_MILLION) * rate.output
+  )
+}

src/menubar-json.ts

@@ -16,6 +16,9 @@ export type PeriodData = {
   outputTokens: number
   cacheReadTokens: number
   cacheWriteTokens: number
+  /// Total Codex credits consumed in the period (issues #408/#495). Optional so
+  /// non-menubar PeriodData producers don't have to compute it.
+  codexCredits?: number
   categories: Array<{ name: string; cost: number; savingsUSD: number; turns: number; editTurns: number; oneShotTurns: number }>
   models: Array<{ name: string; cost: number; savingsUSD: number; calls: number }>
   projects?: Array<{ name: string; cost: number; savingsUSD: number; sessions: number; sessionDetails?: Array<{ cost: number; savingsUSD: number; calls: number; inputTokens: number; outputTokens: number; date: string; models: Array<{ name: string; cost: number; savingsUSD: number }> }> }>
@@ -85,6 +88,8 @@ export type MenubarPayload = {
     inputTokens: number
     outputTokens: number
     cacheHitPercent: number
+    /// Codex credits consumed in the period; 0 when there is no Codex usage.
+    codexCredits: number
     topActivities: Array<{
       name: string
       cost: number
@@ -321,6 +326,7 @@ export function buildMenubarPayload(
       inputTokens: current.inputTokens,
       outputTokens: current.outputTokens,
       cacheHitPercent: cacheHitPercent(current.inputTokens, current.cacheReadTokens),
+      codexCredits: current.codexCredits ?? 0,
       topActivities: buildTopActivities(current.categories),
       topModels: buildTopModels(current.models),
       localModelSavings: breakdowns?.localModelSavings ?? { totalUSD: 0, calls: 0, byModel: [], byProvider: [] },

src/models-report.ts

@@ -1,6 +1,7 @@
 import chalk from 'chalk'
 import stripAnsi from 'strip-ansi'
 
+import { codexCredits } from './codex-credits.js'
 import { formatCost, formatTokens } from './format.js'
 import { getProvider } from './providers/index.js'
 import { CATEGORY_LABELS, type ProjectSummary, type TaskCategory } from './types.js'
@@ -20,6 +21,9 @@ export type ModelReportRow = {
   savingsUSD: number
   savingsBaselineModel: string
   calls: number
+  /// Codex credit consumption (issues #408/#495). null for non-Codex models or
+  /// Codex models without a known credit rate.
+  credits: number | null
   topCategory?: TaskCategory
   topCategoryCost?: number
   topCategoryShare?: number
@@ -156,6 +160,16 @@ export async function aggregateModels(projects: ProjectSummary[], opts: Aggregat
       savingsUSD: bucket.savingsUSD,
       savingsBaselineModel: bucket.savingsBaselineModel,
       calls: bucket.calls,
+      // outputTokens already includes reasoning (folded in above), and for Codex
+      // inputTokens is non-cached with cacheReadTokens holding cached input, which
+      // is exactly what the credit rates expect.
+      credits: bucket.provider === 'codex'
+        ? codexCredits(bucket.model, {
+            inputTokens: bucket.inputTokens,
+            cachedReadTokens: bucket.cacheReadTokens,
+            outputTokens: bucket.outputTokens,
+          })
+        : null,
     }
 
     if (!opts.byTask) {
@@ -553,6 +567,7 @@ export function renderJson(rows: ModelReportRow[]): string {
       costUSD: r.costUSD,
       savingsUSD: r.savingsUSD,
       savingsBaselineModel: r.savingsBaselineModel,
+      credits: r.credits,
     })),
     null,
     2,

src/usage-aggregator.ts

@@ -6,6 +6,7 @@ import { getLocalModelSavingsConfigHash, getShortModelName } from './models.js'
 import { getAllProviders } from './providers/index.js'
 import { aggregateProjectsIntoDays, buildPeriodDataFromDays } from './day-aggregator.js'
 import { aggregateModelEfficiency } from './model-efficiency.js'
+import { aggregateModels } from './models-report.js'
 import { scanAndDetect } from './optimize.js'
 import { getDaysInRange, ensureCacheHydrated, loadDailyCache, emptyCache, BACKFILL_DAYS, toDateString, type DailyCache } from './daily-cache.js'
 
@@ -157,6 +158,15 @@ export async function buildMenubarPayloadForRange(periodInfo: PeriodInfo, opts:
     currentData = buildPeriodData(periodInfo.label, scanProjects)
   }
 
+  // Codex credits for the period. Reuses the models aggregation (folds reasoning
+  // into output, keeps non-cached input + cached-read separate) so the figure
+  // matches the official credit rates.
+  const modelRows = await aggregateModels(scanProjects)
+  currentData.codexCredits = modelRows.reduce(
+    (sum, r) => sum + (r.provider === 'codex' && r.credits != null ? r.credits : 0),
+    0,
+  )
+
   // PROVIDERS
   // For .all: enumerate every provider with cost across the period (from cache) + installed-but-zero.
   // For specific: just this single provider with its scoped cost.

tests/codex-credits.test.ts

@@ -0,0 +1,53 @@
+import { describe, expect, it } from 'vitest'
+import { codexCredits, codexCreditRate } from '../src/codex-credits.js'
+
+describe('codexCreditRate', () => {
+  it('resolves the documented per-model rates', () => {
+    expect(codexCreditRate('gpt-5.5')).toEqual({ input: 125, cachedInput: 12.5, output: 750 })
+    expect(codexCreditRate('gpt-5.4')).toEqual({ input: 62.5, cachedInput: 6.25, output: 375 })
+    expect(codexCreditRate('gpt-5.4-mini')).toEqual({ input: 18.75, cachedInput: 1.875, output: 113 })
+  })
+
+  it('tolerates codex suffix variants and casing', () => {
+    expect(codexCreditRate('GPT-5.5-codex')?.input).toBe(125)
+    expect(codexCreditRate('gpt-5.4-codex-mini')?.input).toBe(18.75)
+  })
+
+  it('returns null for models with no known credit rate', () => {
+    expect(codexCreditRate('gpt-4o')).toBeNull()
+    expect(codexCreditRate('claude-opus-4-8')).toBeNull()
+  })
+})
+
+describe('codexCredits', () => {
+  it('charges 1M input tokens at the input rate', () => {
+    expect(codexCredits('gpt-5.5', { inputTokens: 1_000_000, cachedReadTokens: 0, outputTokens: 0 })).toBe(125)
+  })
+
+  it('charges 1M output tokens at the output rate', () => {
+    expect(codexCredits('gpt-5.5', { inputTokens: 0, cachedReadTokens: 0, outputTokens: 1_000_000 })).toBe(750)
+  })
+
+  it('charges cache-read tokens at the cheaper cached rate', () => {
+    expect(codexCredits('gpt-5.5', { inputTokens: 0, cachedReadTokens: 1_000_000, outputTokens: 0 })).toBe(12.5)
+  })
+
+  it('folds reasoning tokens into the output rate', () => {
+    // 500k output + 500k reasoning = 1M output-billed => 750 credits.
+    expect(codexCredits('gpt-5.5', { inputTokens: 0, cachedReadTokens: 0, outputTokens: 500_000, reasoningTokens: 500_000 })).toBe(750)
+  })
+
+  it('sums a mixed record (gpt-5.4)', () => {
+    // 2M input (125) + 1M cached (6.25) + 0.5M output (187.5) = 318.75
+    const credits = codexCredits('gpt-5.4', { inputTokens: 2_000_000, cachedReadTokens: 1_000_000, outputTokens: 500_000 })
+    expect(credits).toBeCloseTo(125 + 6.25 + 187.5, 6)
+  })
+
+  it('clamps negative / non-finite token counts to 0', () => {
+    expect(codexCredits('gpt-5.5', { inputTokens: -100, cachedReadTokens: NaN, outputTokens: 1_000_000 })).toBe(750)
+  })
+
+  it('returns null for an unknown model', () => {
+    expect(codexCredits('gpt-4o', { inputTokens: 1_000_000, cachedReadTokens: 0, outputTokens: 0 })).toBeNull()
+  })
+})

tests/models-report.test.ts

@@ -125,6 +125,37 @@ describe('aggregateModels', () => {
     expect(claudeRow.totalTokens).toBe(1800 + 300 + 800 + 13000)
   })
 
+  it('computes Codex credits per model and leaves non-Codex / unknown models null', async () => {
+    const rows = await aggregateModels([makeProject([
+      // gpt-5.5: 1M non-cached input (125) + 1M cached read (12.5) + 1M output (750) = 887.5 credits
+      makeTurn('feature', [
+        makeCall({ provider: 'codex', model: 'gpt-5.5', input: 1_000_000, output: 1_000_000, cacheRead: 1_000_000, costUSD: 9 }),
+      ]),
+      // codex but no known credit rate -> null
+      makeTurn('feature', [
+        makeCall({ provider: 'codex', model: 'gpt-5', input: 1000, output: 80, costUSD: 1.2 }),
+      ]),
+      // non-codex provider -> null even if tokens present
+      makeTurn('feature', [
+        makeCall({ provider: 'claude', model: 'claude-sonnet-4-6', input: 1000, output: 200, costUSD: 5 }),
+      ]),
+    ])])
+    const byKey = Object.fromEntries(rows.map(r => [`${r.provider}:${r.model}`, r]))
+    expect(byKey['codex:gpt-5.5']!.credits).toBeCloseTo(887.5, 6)
+    expect(byKey['codex:gpt-5']!.credits).toBeNull()
+    expect(byKey['claude:claude-sonnet-4-6']!.credits).toBeNull()
+  })
+
+  it('includes credits in the JSON output', async () => {
+    const rows = await aggregateModels([makeProject([
+      makeTurn('feature', [
+        makeCall({ provider: 'codex', model: 'gpt-5.5', input: 0, output: 1_000_000, cacheRead: 0, costUSD: 9 }),
+      ]),
+    ])])
+    const parsed = JSON.parse(renderJson(rows))
+    expect(parsed[0].credits).toBeCloseTo(750, 6)
+  })
+
   it('does not double-count cache reads when a provider sets both cache fields', async () => {
     // Providers like codex/mux/codebuff populate cacheReadInputTokens AND
     // cachedInputTokens with the same value (Anthropic vs OpenAI vocabulary for
@@ -231,6 +262,7 @@ describe('renderTable', () => {
       savingsUSD: 0,
       savingsBaselineModel: '',
       calls: 0,
+      credits: null,
       ...partial,
     }
   }

tests/usage-aggregator.test.ts

@@ -42,6 +42,9 @@ describe('buildMenubarPayloadForRange', () => {
     expect(Array.isArray(payload.current.topModels)).toBe(true)
     expect(Array.isArray(payload.history.daily)).toBe(true)
     expect(payload.current.retryTax.totalUSD).toBeGreaterThanOrEqual(0)
+    // Codex credits are always present in the payload (display gates them); 0 with no data.
+    expect(typeof payload.current.codexCredits).toBe('number')
+    expect(payload.current.codexCredits).toBeGreaterThanOrEqual(0)
     // optimize:false => scanAndDetect skipped => empty optimize block regardless of data
     expect(payload.optimize).toEqual({ findingCount: 0, savingsUSD: 0, topFindings: [] })
   })