AI Stats is built around a few simple principles that keep the platform reliable, open, and easy to use.
Openness and transparency
- Keep the Gateway and data model auditable and easy to understand.
- Make pricing, routing, and performance visible wherever possible.
One API for every model
- Provide a consistent surface that works across providers and modalities.
- Minimize provider-specific rewrites when models change.
Portability and resilience
- Route requests to healthy providers and recover from upstream issues.
- Avoid lock-in by keeping the API compatible with major model ecosystems.
Observability by default
- Capture usage, latency, and error signals so teams can operate safely.
- Expose diagnostics via the API and dashboard for production visibility.
- Encourage constructive feedback and contributions from all users.
- Maintain clear contributor expectations and a shared code of conduct.
Next steps
Last modified on February 11, 2026