Cost Breakdown

Per-Model Cost Breakdown

Drill into exactly which models cost what, with sortable tables, provider filtering, and cost-per-million calculations.

Model-level granularity

See every model across every provider in one unified, sortable table. Each row shows input tokens, output tokens, cached tokens, credits consumed, request count, and total cost per model. Sort by any column to instantly find your most expensive models, highest-volume models, or best cost-efficiency targets.

Cost per million tokens

An automatically calculated cost efficiency metric that normalizes your spending. Compare how much you actually pay per 1M tokens across different models — accounting for real-world usage patterns including caching, retries, and varying prompt lengths. This metric reveals which models deliver the best value for your specific workloads.

Search and filter

Filter the breakdown table by provider to focus on a single vendor's models, or use full-text search to find any model by name. Combining filters lets you answer precise questions like 'How much did we spend on Claude 3.5 Sonnet this month?' or 'What are our most expensive OpenAI models?' in seconds.

CSV export

Export the full breakdown as a CSV file for further analysis in spreadsheets, BI tools, or internal reporting systems. The export includes every column — model name, provider, input tokens, output tokens, cached tokens, credits, requests, total cost, and cost per million tokens — for any selected time range. Available on the Pro plan.