Pydantic Services Inc. ("PSI") uses certain Subprocessors to assist in providing the Services (including Pydantic Logfire and the Pydantic AI Gateway). Subprocessors are PSI's authorized vendors and third-party service providers that Process Customer Personal Data. Capitalized terms not defined on this page have the meanings ascribed to them by the terms set forth in the Data Processing Addendum or other separate data processing agreement between Customer and PSI, as applicable.
What is a Sub-processor?
When PSI engages third-party service providers in its capacity as a data processor for its Customers' personal data, the General Data Protection Regulation ("GDPR") and a number of other global privacy frameworks call these third-party service providers sub-processors. Sub-processors are service providers who have or potentially will have access to or process personal data that PSI processes for, and on behalf of, its Customers.
This page outlines the types of service providers and Sub-processors we utilize, where they are located, and a description of the work they carry out.
Subscribe to Updates
PSI will give at least fourteen (14) days' notice before adding or replacing a subprocessor that will Process Customer Personal Data, except where a shorter period is reasonably necessary for urgent security or legal reasons. If you object to the change on reasonable data protection grounds, notify PSI in writing within fourteen (14) days of the notice, as described in Section 5 of the Data Processing Addendum. To receive email notifications when this list is updated, subscribe below:
Subprocessors
PSI may engage the following entities to process Customer Personal Data that you include in your use of the Logfire Service:
Effective as of: November 10, 2026
AI providers
Service inference covers prompts and model inputs/outputs processed through built-in Pydantic AI Gateway providers or Logfire AI features. A provider receives this data when a request is routed to it.
Internal use covers Pydantic's business operations, data analysis, and software development. Pydantic's AI policy requires model training to be disabled for all internal use.
| Subprocessor | Service inference | Internal use | Processing location |
|---|---|---|---|
| Anthropic, PBC | Built-in AI Gateway provider | Business productivity, data analysis, and coding (Claude and Claude Code) | US (storage); global (Gateway inference) |
| Google Cloud Platform (Vertex AI) | Built-in AI Gateway provider and Logfire AI features | Not listed | US, EU (Logfire AI); global (Gateway) |
| OpenAI | Built-in AI Gateway provider and Logfire AI features | Business productivity and coding (ChatGPT and Codex) | US; global (Gateway) |
Other subprocessors
| Subprocessor | Purpose | Processing location |
|---|---|---|
| Amazon Web Services | Cloud hosting and email delivery | US, EU |
| Clay, Inc. | Sales enrichment and account management | US |
| Cloudflare, Inc. | Content delivery and traffic routing | US |
| Customer.io, Inc. | Marketing and lifecycle emails | EU |
| Day AI, Inc. | Customer relationship management and interaction history | US |
| Google Cloud Platform | Cloud hosting | US, EU |
| Google LLC | Email, documents, video conferencing, and AI assistance (Gemini) | US |
| HubSpot, Inc. | Customer relationship management | EU (hosting); US (additional processing) |
| Metronome Holdings Inc. | Usage-based billing (Gateway enabled) | US |
| PagerDuty, Inc. | Incident management and alerting | EU |
| PostHog Inc. | Product analytics and submitted sales inquiry contact data | EU |
| Pylon, Inc. | Customer support | US |
| Devenue Inc. (Reo.Dev) | Developer intent and account intelligence | US (hosting) |
| Slack Technologies, LLC | Messaging and file sharing | US |
| Stripe, Inc. | Payment processing | US |