API Reference
Projects Setup
Last updated May 28, 2026
How projects fit into public Cencori integrations.
Overview
Cencori project creation and project settings are managed in the dashboard today. They are not part of the public https://api.cencori.com/v1 runtime API.
For product integrations, the setup flow is:
- Create a project in the dashboard.
- Generate a project API key from the project's API Keys page.
- Configure provider access in the project's Providers page.
- Use that project key from your server when calling the runtime APIs.
Runtime APIs
Project API keys authenticate these public runtime surfaces:
| Use case | Endpoint |
|---|---|
| Native Cencori SDK / REST chat | https://cencori.com/api/ai/chat |
| OpenAI-compatible chat | https://api.cencori.com/v1/chat/completions |
| Models | https://api.cencori.com/v1/models |
| Metrics | https://api.cencori.com/v1/metrics |
Programmatic Project Management
Public project-management endpoints are not available yet. Do not configure product code to call /organizations/{orgSlug}/projects on api.cencori.com.
Dashboard routes under /api/projects/... are session-authenticated and are intended for the signed-in Cencori dashboard, not for runtime product integrations.
Common Setup Mistake
A valid project API key only proves your app can authenticate to Cencori. AI requests still need provider access configured for the project, either through a connected provider, Cencori-managed access, or BYOK.
If /v1/models works but /v1/chat/completions fails with a provider configuration error, keep the same Cencori key and finish provider setup in the dashboard.