> For the complete documentation index, see [llms.txt](https://docs.otari.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.otari.ai/privacy/zdr.md).

# Zero Data Retention

Otari connects to multiple LLM providers to process inference requests. Each provider has its own policies regarding how request and response data are handled, stored, and used. Understanding these policies helps you select providers that meet your application's privacy, security, and compliance requirements.

Zero Data Retention (ZDR) means that a provider does not persist prompts or model-generated responses after processing a request. Data is processed temporarily in memory for as long as it takes to generate the response.

When evaluating a provider's data handling policies, there are three key aspects:

| Concept              | Description                                                                              |
| -------------------- | ---------------------------------------------------------------------------------------- |
| **Data retention**   | Whether prompts and responses are stored after processing, and for how long.             |
| **Model training**   | Whether request and response data can be used to train or improve models.                |
| **Metadata logging** | Whether operational data, such as timestamps, token usage, and request IDs, is retained. |

## **Provider data retention policies**

The following table summarizes the retention classifications of providers supported by Otari.

| Provider       | Data retention | Model training | Privacy Policy                                                                       | Models |
| -------------- | -------------- | -------------- | ------------------------------------------------------------------------------------ | ------ |
| Amazon Bedrock | Zero retention | No             | [Privacy](https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html) | 49     |
| Fireworks AI   | Zero retention | No             | [Privacy](https://fireworks.ai/privacy-policy)                                       | 15     |
| Mistral AI     | Zero retention | No             | [Privacy](https://legal.mistral.ai/terms/get-started/#privacy-policy)                | 25     |
| Nebius         | Zero retention | No             | [Privacy](https://docs.nebius.com/legal/privacy)                                     | 22     |
| OVH Cloud      | Zero retention | No             | [Privacy](https://us.ovhcloud.com/legal/privacy-policy/)                             | 10     |
| Together AI    | Zero retention | No             | [Privacy](https://www.together.ai/privacy)                                           | 14     |

Regarding metadata logging, note that even when Zero Data Retention is enabled, providers may retain operational metadata for monitoring, security, and compliance. This data may include timestamps, token counts, request identifiers, and usage statistics, but should not include prompt or response content under the applicable ZDR policy.

For more information about the use of third parties, please refer to our [Privacy Policy](https://otari.ai/privacy).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.otari.ai/privacy/zdr.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.