Security & Trust

If you handle client-sensitive work, this is the short version of how we handle your data.

Core Commitments

No AI model training on your data
Uploads and outputs are used only to generate your result, not to train models.
0 data retention for source files and outputs on our side
We do not keep long-term archives of your uploaded files or generated outputs.
Auto-routed inference
Jobs are routed across providers automatically, and we apply deletion and minimization settings where they are available.
Clear third-party handling
Routed providers may process data briefly during generation. We keep that window as small as possible and prioritize providers that support strict cleanup.
Encrypted end to end
Data is protected in transit (TLS) and at rest.
SOC 2 compliant infrastructure foundation
ModelMyImage runs on infrastructure that supports SOC 2 compliant controls.

If your team needs a security review, use this form: Request Security Review