Provider Comparison (preview)

Your real usage volume, priced against each provider's currently published rates. Anthropic uses your actual model mix; for OpenAI and Google, each Claude model is mapped to the closest comparable model on that provider and the same token volume is priced against it. Those rates are seeded manually, not pulled live.

Local models are excluded from every column: they cost nothing per token, so re-pricing them against a paid API would invent a bill that nobody would ever pay.