BackLLM gateway finance audit

See profit as the space between revenue and supplier cost.

Meter input, output, and cached tokens once. Let amount stay as customer revenue, then add analytics-only metrics for total revenue and cost so finance can chart margin without double charging customers.

revenue

$20,350

total_tokens.amount

cost

$5,899

cost.amount

gross margin

71.0%

computed after query

POST /analysis/query · grouped series

Revenue over cost, profit in between

$14,451 gross profit

Revenue

Cost

Gross profit band

Metric abstraction

No finance-only fields. Just rated metrics.

The customer-facing token metrics keep their quantity and amount. Analytics-only metrics reuse those results for reporting, without entering debit.

metricquantityamountdebit

input_tokens$.usage.input_tokensCustomer revenueyes

output_tokens$.usage.output_tokensCustomer revenueyes

cached_tokens$.usage.input_tokens_details.cached_tokensCustomer revenueyes

total_tokensitems[input/output/cached]Revenue rollupanalytics_only

cost0Supplier price bookanalytics_only

Rule

Define metrics once

Token metrics carry customer amount. Virtual metrics reuse those facts for revenue rollup and supplier cost.

Event

Report price hints with the request

The gateway sends customer prices and cost prices in the same event, keyed by metric.

Query

Draw from analysis/query

The chart calls the public analysis query API for total_tokens and cost, then renders grouped series directly.

ExtractorRuleSet

Create and publish through /v1/projects/:project_id/extractor-rule-sets.

demo

{
  "id": "ers_llm_gateway_finance",
  "name": "LLM Gateway Finance Metrics",
  "subject": {
    "default_type": "api_key",
    "type_path": "$.meta.subject_type",
    "id_path": "$.meta.subject_id",
    "required": true
  },
  "dimensions": [
    { "name": "provider", "path": "$.provider", "type": "string", "required": true },
    { "name": "model", "path": "$.model", "type": "string", "required": true },
    { "name": "customer_id", "path": "$.meta.customer_id", "type": "string", "required": true }
  ],
  "charge_item_extractors": [
    {
      "name": "input-tokens",
      "metric": "input_tokens",
      "path": "$.usage.input_tokens",
      "type": "number",
      "unit": "token",
      "skip_if_missing": true,
      "currency": "USD",
      "unit_price_expr": "dec('0')",
      "amount_expr": "dec_mul(dec_div(dec(item['quantity']), dec(item['pricing_unit'])), dec(item['unit_price']))"
    },
    {
      "name": "output-tokens",
      "metric": "output_tokens",
      "path": "$.usage.output_tokens",
      "type": "number",
      "unit": "token",
      "skip_if_missing": true,
      "currency": "USD",
      "unit_price_expr": "dec('0')",
      "amount_expr": "dec_mul(dec_div(dec(item['quantity']), dec(item['pricing_unit'])), dec(item['unit_price']))"
    },
    {
      "name": "cached-tokens",
      "metric": "cached_tokens",
      "path": "$.usage.input_tokens_details.cached_tokens",
      "type": "number",
      "unit": "token",
      "skip_if_missing": true,
      "currency": "USD",
      "unit_price_expr": "dec('0')",
      "amount_expr": "dec_mul(dec_div(dec(item['quantity']), dec(item['pricing_unit'])), dec(item['unit_price']))"
    },
    {
      "name": "total-tokens",
      "metric": "total_tokens",
      "quantity_expr": "dec_add(dec_add(dec(items['input_tokens']['quantity']), dec(items['output_tokens']['quantity'])), dec(items['cached_tokens']['quantity']))",
      "type": "number",
      "unit": "token",
      "currency": "USD",
      "amount_expr": "dec_add(dec_add(dec(items['input_tokens']['amount']), dec(items['output_tokens']['amount'])), dec(items['cached_tokens']['amount']))",
      "control": { "analytics_only": "true" }
    },
    {
      "name": "cost",
      "metric": "cost",
      "quantity_expr": "dec('0')",
      "type": "number",
      "unit": "event",
      "currency": "USD",
      "amount_expr": "dec_add(dec_add(dec_mul(dec_div(dec(items['input_tokens']['quantity']), dec(pricing_hints['cost.input_tokens']['pricing_unit'])), dec(pricing_hints['cost.input_tokens']['unit_price'])), dec_mul(dec_div(dec(items['output_tokens']['quantity']), dec(pricing_hints['cost.output_tokens']['pricing_unit'])), dec(pricing_hints['cost.output_tokens']['unit_price']))), dec_mul(dec_div(dec(items['cached_tokens']['quantity']), dec(pricing_hints['cost.cached_tokens']['pricing_unit'])), dec(pricing_hints['cost.cached_tokens']['unit_price'])))",
      "control": { "analytics_only": "true" }
    }
  ]
}

Profit analysis

The drawing layer stays simple.

The API returns metric buckets. The chart joins matching buckets, calculates profit and margin, then renders revenue and cost as two independent lines.

Start with your gateway data

revenue querymetrics: ['total_tokens'], measures: ['amount']

cost querymetrics: ['cost'], measures: ['amount']

bucket joingross_profit = total_tokens.amount - cost.amount

margingross_margin = gross_profit / total_tokens.amount