Cost Tracking

Monitor and control LLM spending across providers with built-in cost tracking, budget management, and model pricing data for 100+ models.

Setup

Wrap any LLMClient with CostAwareLLMClient to start tracking costs automatically. Every request records its token usage and calculates cost based on the model's pricing.

// Wrap with default cost tracking
LLMClient tracked = CostAwareLLMClient.wrap(client);

// Or use the builder for full control
CostAwareLLMClient tracked = CostAwareLLMClient.builder()
    .client(client)
    .costTracker(new InMemoryCostTracker())
    .budgetManager(budget)
    .build();

Querying Costs

Access accumulated cost data at any time. You can get the total across all models, break down by individual model, or query a specific time range.

tracker.getTotalCost();                            // Total across all models
tracker.getCostByModel("gpt-4o");                  // Per-model breakdown
tracker.getCostInRange(startTime, endTime);         // Time-range query

Set spending limits to prevent runaway costs. When the budget is exceeded, the CostAwareLLMClient will reject further requests. You can also set an alert threshold to get early warnings before hitting the limit.

BudgetManager budget = BudgetManager.builder()
    .limit(new BigDecimal("50.00"))
    .daily()                          // Duration.ofDays(1)
    .alertThreshold(0.80)
    .build();

// Or monthly budget
BudgetManager monthly = BudgetManager.builder()
    .limit(new BigDecimal("500.00"))
    .monthly()                        // Duration.ofDays(30)
    .alertThreshold(0.80)
    .build();

Model Pricing

TnsAI ships with built-in pricing data for 100+ models so cost calculations work out of the box. Prices are in USD per 1 million tokens.

Model	Input	Output
gpt-4o	$2.50	$10.00
gpt-4o-mini	$0.15	$0.60
claude-sonnet-4	$3.00	$15.00
claude-opus-4	$15.00	$75.00
claude-3.5-haiku	$0.80	$4.00
gemini-2.5-pro	$1.25	$10.00
gemini-2.5-flash	$0.15	$0.60
gemini-2.0-flash	$0.075	$0.30

Programmatic Usage

Look up pricing for any model and calculate costs for a given token count.

ModelPricing pricing = ModelPricing.forModel("gpt-4o");
BigDecimal inputCost = pricing.calculateInputCost(1000);   // 1000 tokens
BigDecimal outputCost = pricing.calculateOutputCost(500);

Usage Records

If you need to record usage manually (for example, from external API calls), you can create UsageRecord objects and pass them to the tracker directly.

tracker.record(UsageRecord.builder()
    .modelId("gpt-4o")
    .inputTokens(1000)
    .outputTokens(500)
    .build());