Data retention
Data Retention Policy
Prompt content is not persisted. The usage metadata that billing needs is retained only as long as necessary, and this policy explains its purpose and scope.
Scope
This policy supplements our Privacy Policy and describes how data is retained for the inference API at b300.powerchampion.ai.
Prompt and output content
Prompt content and model output are processed only to complete the inference request. They are released when request processing finishes and are not written to persistent storage. Conversation content is not kept in billing records or operational usage logs.
Usage metadata
We retain metering metadata such as timestamps, API-key identifiers, model identifiers, measured token or media usage, a billing record identifier, and an operational status note when needed, only as long as needed for billing and reconciliation purposes. These records do not contain prompt or output content.
No training use
Prompt content and model output are not used to train, fine-tune, or evaluate machine-learning models.
Deletion and legal obligations
When records are no longer needed for their stated billing, accounting, security, or legal purpose, they are deleted or de-identified. If applicable law requires a record to be preserved, we keep only the minimum required scope for the required period.
Contact
Power Champion Investment Limited — info@powerchampion.org