Sub-Processors
Last updated: September 30, 2026
High Agency, Inc. uses the following sub-processors to provide the pen.dev application and related services. A sub-processor is a third party that processes data on our behalf. This list is updated prior to engaging any new sub-processor.
AI & Content Services (Backend-Routed)
| Name | Purpose |
|---|---|
| Fireworks AI | AI model hosting (pen.dev Pro models) |
| Image generation (Gemini) | |
| OpenAI | Image generation |
| Portkey | AI gateway (pen.dev Pro models) |
| Quiver AI | SVG generation |
| Recraft | Image editing & vectorization |
| Unsplash | Stock image search |
Infrastructure & Operations
| Name | Purpose |
|---|---|
| Google Cloud Platform | Cloud infrastructure, file storage & backend services |
| Vercel | Application hosting & backend infrastructure |
| Neon | Database hosting |
| Stripe | Payment processing |
| SendGrid | Transactional & product email |
| PostHog | Product analytics |
| GoatCounter | Web analytics (cookie-free) |
| Sentry | Error monitoring |
| Slack | Internal communication |
| Google Workspace | Business email & documents |
Note on device-direct AI providers: Text-based AI features that run on the user's own provider account or API key (for example Anthropic, OpenAI and Google Gemini text models, GitHub Copilot, or any other model provider or custom endpoint the user adds) connect directly from the user's device using the user's own credentials and do not route through High Agency infrastructure. These providers are therefore not sub-processors of High Agency, Inc. This does not apply to pen.dev Pro models, which are provided by High Agency and routed through the inference partners listed above.
If you have questions about our sub-processors or data processing practices, contact us at hq@pen.dev.