Request Routing

RouteAPI lets business systems call a single unified API entry point. The platform handles service selection, availability checks, and request delivery in the background.

What you need to configure

Most businesses only need to focus on:

• The RouteAPI model ID used in the request.
• Whether streaming output is enabled.
• Whether optional capabilities such as tool calls, vision understanding, or structured output are required.
• API Key and usage limits.

The platform uses this information to send the request to a suitable model service.

Request handling flow

1. RouteAPI validates the API Key and account permissions.
2. RouteAPI reads the model ID in the request.
3. RouteAPI determines whether the request needs special capabilities.
4. RouteAPI sends the request through an available service path.
5. RouteAPI returns the result using the protocol expected by the client.

Your business code does not need to expose internal service details.

Usage recommendations

• Keep model IDs stable in business code.
• Confirm supported capabilities in the model list in the console.
• Test important flows in advance using the same request format as production.
• When troubleshooting latency, errors, or costs, check usage logs first.