Privacy Policy

Effective date: April 7, 2026

TokenDock provides OpenAI-compatible access to a selected set of AI models. This policy explains how we handle account data, billing data, usage metadata, and inference request data when you use TokenDock.

1. Customer data ownership

Your prompts, inputs, uploaded content, and model responses are your customer data. They belong solely to you. TokenDock does not claim ownership of your prompt or response data.

2. No prompt storage, monitoring, training, or sale

TokenDock does not store user prompts or model responses as part of normal service operation. We do not monitor prompt content in transit. We do not use prompt data to train nor fine-tune AI models. We do not sell prompt or customer data.

3. Inference request processing

We process request content only transiently to route your request, call the selected model provider, and return the response to you. Request and response content is treated as confidential and is not retained by TokenDock after the request completes.

TokenDock may store operational metadata such as request ID, timestamp, account or API key reference, model ID, token counts, cost, status code, and latency for billing, abuse prevention, reliability, and support. Operational metadata does not include the full prompt or completion content.

4. Account and billing data

We store the account, API key, wallet, payment, usage, and audit information needed to provide prepaid billing, invoice support, security review, and customer support. Raw API keys are shown only at creation time and are not stored in plaintext by TokenDock.

5. Model providers

TokenDock uses a selected number of model providers to generate responses for supported models. If you want to review a specific provider's terms or privacy policy, contact us at [email protected].

6. Supported regions

TokenDock currently supports model routing in the United States, Singapore, and Germany. If you have a specific data residency requirement, contact support before sending production traffic.

7. Security

TokenDock uses TLS for data in transit and access controls for service administration. We separate public runtime access from administrative access and keep operational secrets out of source control.

8. Data retention

We retain account, billing, audit, and usage metadata as needed to operate TokenDock, resolve disputes, meet legal obligations, and protect the service. We do not retain prompt or response content in normal service operation.

9. Your responsibilities

Do not send secrets, regulated personal data, or confidential third-party data unless you have the right to process it through TokenDock and the selected upstream provider.

10. Contact

If you have questions about this privacy policy, contact us at [email protected].