Zero-config pricingPricing for 2,600+ models is resolved automatically from the LiteLLM community database, cached daily.
Drop-in traitReplace Promptable with TracksAiCost on your agent — every prompt() call is tracked transparently.
Works everywhere laravel/ai worksAnthropic, OpenAI, Gemini, Mistral, DeepSeek, Groq — provider auto-detection from the model name.
Local overridesPin custom prices or override LiteLLM via the published config. Wildcard prefix matching included.
Readonly DTOGet back a clean AiCostResult with input/output USD costs, token counts, model, and provider.
Per-agent totalsRoll up costs across an agent's prompt() calls with totalCostUsd() and costs() — ready for budgets, logs, and dashboards.