Developer Migration Trigger
- Same request/response structure
- Minimal or no code changes
- Full control over model routing
- No vendor lock-in
CONTROL PLANE
The API and Control Layer for Your Private AI Environment
SecludedAI Gateway sits between users, applications, and your private AI execution engine. It handles authentication, routing, model abstraction, and request management.
Replace OpenAI Without Rewriting Your App
SecludedAI Gateway exposes an OpenAI-compatible API so you can route existing AI calls through infrastructure you control.
Strong Fit For
- Private apps and internal tools
- SaaS products reducing provider dependency
- Teams migrating existing SDK-based workflows
Gateway Responsibilities
Gateway is where the system becomes operational. It provides the secure API layer that applications and interfaces use to communicate with your private AI infrastructure.
Gateway handles user authentication, API requests, request tracking, model selection logic, and communication with SecludedAI Core. It also creates a stable contract between the user experience layer and the underlying model execution layer.
By separating Gateway from Core, SecludedAI gains a cleaner architecture that is easier to manage, secure, and extend over time.
Key Points
- User authentication
- API routing and request handling
- Model abstraction layer
- Job tracking and response flow
- Control-plane role between Studio and Core
Use It Like OpenAI. Control It Like Infrastructure.
SecludedAI Gateway is designed to support an OpenAI-compatible API layer, allowing existing applications, tools, and workflows to connect with minimal changes.
Instead of sending requests to a third-party provider, your applications can route requests through your own Gateway, where you control authentication, model routing, and execution.
SecludedAI Gateway can expose an OpenAI-compatible API, so you can keep your existing integrations while moving the intelligence layer under your control.
Built for Migration, Not Lock-In
Most AI systems today are tightly coupled to external providers. SecludedAI Gateway allows you to decouple your applications from that dependency over time. You can start with familiar API patterns and evolve toward a fully controlled AI stack without rewriting your entire system.
This makes it possible to
- Replace external AI API calls with your own infrastructure
- Maintain compatibility with existing tools and SDKs
- Control how models are selected, routed, and executed
- Keep request handling inside your environment
Supporting Points
- OpenAI-style request/response structure
- Drop-in replacement for many existing integrations
- Compatible with common SDK patterns
- Centralized control over model behavior
SecludedAI Gateway FAQ
What does SecludedAI Gateway do?
Gateway is the control plane for the SecludedAI platform. It handles authentication, API requests, request routing, model abstraction, and communication with Core.
Why is Gateway separate from Core?
Gateway and Core have different responsibilities. Gateway manages access, request flow, and API contracts. Core focuses on private model execution. Separating them creates a cleaner and more secure architecture.
Can Gateway support different model mappings?
Yes. Gateway is designed as an abstraction layer so it can expose logical models to client applications while mapping them to the actual runtimes and model choices available in the customer environment.
Is Gateway public-facing?
Gateway is customer-owned, but SecludedAI Tunnel is intended to be the managed way hosted Studio reaches it. The goal is to avoid forcing customers to expose Gateway to the public internet in the traditional sense.
Does Gateway store user accounts?
Gateway is designed to handle authentication and request-layer responsibilities within the customer environment. Specific storage and configuration details depend on the deployment, but Gateway serves as the access and control layer.
Can Gateway power other apps besides Studio?
Yes. Gateway is intended to act as the API/control layer for the customer's AI environment, which means it can support additional private apps and integrations over time. With an OpenAI-compatible API layer, many existing tools and SDK-based workflows can connect through Gateway while model routing and execution remain under customer control.