🛡️ What is Zero Data Retention?

ZDR-compliant endpoints process your AI requests without storing any data. This means:

• No prompt logging or training data usage
• No conversation history retention
• Maximum privacy for sensitive data
• Ideal for confidential business information, personal data, or sensitive research

Note on Caching: In-memory caching of prompts is typically not considered "retaining" data (or it is at least of a very different nature than organizations that log prompts, write them to databases, or even use them for additional training or publish prompts). This keeps repeated prompt data in an in-memory cache in the provider's datacenter so that the repeated part of the prompt does not need to be re-processed, leading to considerable cost savings. OpenRouter's stance is that in-memory caching is compatible with ZDR, and therefore allows endpoints/models with implicit caching to be hit when a ZDR routing policy is in effect. If you need complete non-retention (no caching), use the "No Caching" filter.

Read OpenRouter's official ZDR policy →

⚠️ Disclaimer

I have not personally audited this list. This catalog relies on OpenRouter's research and public documentation. While I strive to maintain accurate information, always verify ZDR compliance directly with providers before handling sensitive data.

🌍 About OpenCode Go

OpenCode Go uses ZDR policy providers. This plan is designed primarily for international users, with models hosted in the US, EU, and Singapore for stable global access. Our providers follow a zero-retention policy and do not use your data for model training. See the privacy section →