Legal

Data Policy

Last updated August 7, 2026

This Data Policy describes how Oh See Arr handles Customer Content submitted to our OCR and inference APIs. It is designed to answer the same questions developers ask of other inference providers: what is retained, whether data is used for training, and what metadata we keep to run the product.

Account, billing, analytics, and crash data about the Service itself are covered by the Privacy Policy. Use of the Service remains subject to the Terms of Service. If this Data Policy conflicts with the Privacy Policy regarding Customer Content retention or training, this Data Policy controls.

1. Two kinds of data

Like other API platforms, we distinguish:

Usage metadata is always collected as needed to authenticate, meter, bill, secure, and operate the Service. Customer Content is handled as described below.

2. No training on Customer Content

We do not use Customer Content to train, fine-tune, or otherwise improve machine learning models — ours or anyone else's. Your documents are processed to fulfill your request, not to build datasets or model weights. There is no opt-in training program for API Customer Content.

3. Default: no retention of Customer Content

By default, we do not retain Customer Content after a request is completed. Typical request flow:

  1. Authenticate the request (API key).
  2. Receive Inputs over encrypted transport (TLS).
  3. Process Inputs on compute infrastructure solely to generate Outputs.
  4. Return Outputs to you.
  5. Discard Inputs and Outputs from our systems when processing finishes.

We may keep usage metadata associated with the request without storing the underlying document or Output text.

4. Summary

Endpoint / useUsed for trainingCustomer Content retention
/v1/chat/completions (OCR inference)NoNone by default; discarded after the request completes
/v1/modelsN/ANo Customer Content involved
Debug retention (opt-in only)NoOnly with your express permission; limited duration for troubleshooting

5. Optional retention for debugging

If you expressly permit it — for example by opting in when reporting a bug, sharing a failing request with support, or enabling a debug retention control we offer — we may retain specific Customer Content for a limited time solely to diagnose and fix issues affecting your workload.

Debug-retained Customer Content will not be used for training. We delete it when the debugging purpose ends, or earlier upon your request when feasible. We may also retain information when required by law.

6. Security and subprocessors

Customer Content is transmitted using industry-standard encryption in transit (TLS). It may be processed on infrastructure operated by us or our hosting and compute providers solely to deliver the Service. Providers that process Customer Content on our behalf are not authorized to use it for their own training or product improvement.

Processing generally occurs in the United States. If your organization needs a Data Processing Addendum, subprocessor list, or specific contractual terms, contact support@ohseearr.com before sending regulated data.

7. Your responsibilities

You are responsible for having the rights and notices needed to submit Customer Content to the Service, including personal data of your end users. Do not submit material you are legally prohibited from processing through a third-party API. You remain responsible for how you use Outputs.

8. Changes

We may update this Data Policy from time to time. We will post the updated policy with a new "Last updated" date. If we materially change how we retain or use Customer Content — for example by introducing default retention or training — we will provide prominent notice before those changes apply.

9. Contact

Questions about this Data Policy: support@ohseearr.com.

Related: Terms of Service · Privacy Policy