This is the most useful message you can send. Please include:
We verify against the provider before publishing a change, and the correction ships with the test suite re-run.
Tell us the model and provider. We add models that developers are actively shipping with, prioritised by how often they come up in cost discussions. If a provider publishes a batch tier, cached rate or context surcharge, we model it — a row that ignores those would be misleading.
If a calculation looks wrong, please send your shareable link (the "Copy shareable link" button encodes every input in the URL). That reproduces your exact figures and removes any ambiguity about what you entered.
Expect a reply within a few days. We do not run a support desk, and we do not sell, rent or share your address.
We maintain a cost calculator. We cannot help with your OpenAI or Anthropic account, billing disputes, API keys, or rate limits — please raise those with the provider directly. We also do not offer consulting or implementation work through this address.