Data Processing & Retention

This section explains what happens to the data you submit to Pixellect ("Pixellect", "we", "us"), operated by Tailored AI LLC, 30 N Gould St Ste N, Sheridan, WY 82801, USA. It is written so you can understand exactly where your images and text go and how long they are kept.

1. How your data flows

Pixellect is a forwarding service. When you use the editor, the images and text (prompts) you submit are sent directly to the third-party AI providers listed below, which run the models that generate your results. The results are returned to you through our service.

We do not store your input images, prompts, generated outputs, or chat transcripts on our own systems. We hold your data only transiently in memory for the moment it takes to pass your request to the relevant provider and return the result to you. We do not keep copies, and we do not use your content to train any model.

Because the models run on third-party infrastructure, your data is retained — briefly — by those providers according to their own policies. Those providers, their roles, and their retention behavior are described below. We select the specific models offered in Pixellect and may change them from time to time; regardless of which model handles your request, the provider-level retention behavior described here applies.

2. Who processes your data (subprocessors)

All four providers are based in the United States and may process and store data on servers in the United States and other countries.

3. Provider retention details

OpenAI

OpenAI retains API inputs and outputs for up to 30 days to operate the service and detect abuse, after which they are deleted, unless it is legally required to keep them longer. OpenAI does not use API data to train its models by default.

OpenRouter

Requests routed through OpenRouter are not retained: OpenRouter does not store prompts unless prompt logging is enabled, and we do not enable prompt logging. We additionally send every request with Zero Data Retention enforced (zdr: true) and data collection denied (data_collection: "deny"). This restricts routing to downstream inference hosts that do not retain your prompts and do not train on them. Your content is processed for inference and not stored afterward. This configuration applies to every model we route through OpenRouter, including any models we add or change in future.

Replicate

For requests made through Replicate's API, all input parameters, output values, output files, and logs are automatically removed approximately one hour after processing. Replicate does not use this data to train models by default. During that short window, generated output files are served from public delivery URLs.

Fal.ai

Fal stores two kinds of data separately: the request payload (the JSON of your inputs and outputs) and the generated media files.

We send every request with X-Fal-Store-IO: 0, which means Fal does not store the request payloads (your input and output JSON).

Generated media files (the images produced by the model) are still uploaded to Fal's content delivery network (CDN) and stored there for a limited period. By default these files are served from public URLs — anyone who has the URL can access the file until it expires.

4. Retention we cannot switch off

Regardless of the settings above, some retention is required by law or by the providers' legal obligations and cannot be disabled:

• Illegal content. Content that a provider's systems flag as apparent child sexual abuse material (CSAM) may be retained and reported to authorities as required by law, even where zero-retention settings are otherwise in effect. This applies to image content and is relevant to any image you submit.
• Legal holds and abuse prevention. A provider may retain data for longer than the periods above where retention is required by law, by a court order, or is reasonably necessary to protect the provider, its users, or the public from harm.

We do not control these exceptions, and they apply at the provider level.

5. Data we do keep

We do not store your content, but to operate the service and your subscription we do process limited account and billing data (for example, your email address, subscription status, and payment records handled by our payment processor). This data is retained for as long as your account is active and as required for tax, accounting, and legal purposes.

6. International transfers

Pixellect is available to users outside the United States. The AI providers described above process data in the United States and potentially other countries. If you are located outside the United States — including in the EU/EEA or the UK — the images and text you submit are transferred to the United States for processing.

For these transfers, we rely on the European Commission's Standard Contractual Clauses (SCCs), together with the UK International Data Transfer Addendum for data subject to UK law, as the legal transfer mechanism. Each of the AI providers we use publishes a Data Processing Agreement (DPA) that names it as a data processor and relies on the SCCs (and, where applicable, the UK Addendum) for international transfers. Where a provider uses its own subprocessors outside the EEA, those onward transfers are likewise covered by SCCs or an applicable adequacy decision.

Our providers run inference on US-based infrastructure, so default processing is not located in the EU; the SCCs are the safeguard that covers that transfer.

7. Your choices and rights

Because we do not retain your content, requests to access or delete submitted images and prompts are, in practice, satisfied by the automatic deletion described above. For account and billing data, you may contact us at pixellect@tailoredpod.ai to request access or deletion.

To exercise any right, or for questions about this section, contact us at pixellect@tailoredpod.ai.

8. Changes

Provider policies, the models we offer, and our configuration may change. We will update this section and revise the "last updated" date when they do.