Choose models and lines
Check model IDs, protocols, capabilities, line pricing, and key-specific availability.
Select a task and protocol
Open the model catalog and choose a text, image, video, or audio model. Check the complete model ID, endpoint types, capabilities, and billing unit on its detail page. Put the ID, not the display name, in model.
Chat Completions, Responses, Messages, and Gemini native APIs use different request bodies. Choose a model and line supporting the protocol required by your client. Changing the URL alone does not establish field compatibility.
Choose a line
The same model may have different prices and availability across lines. Before creating a key or switching its line, check model access and whether your balance or subscription applies. Token prices use USD / 1M tokens; per-request models show their own billing unit.
A model name does not establish its context limit, image input, tools, or reasoning support. Validate the capabilities your production application needs.
Check key-specific availability
The public catalog helps with selection. GET https://gateway.irouter.io/v1/models shows models visible to the current key. Keys and lines may return different lists. This response is not a complete capability or price table.
Query the list and replace <model-id> in examples with an actual ID. If the catalog or response is empty, check the selected line and permissions or ask an administrator to confirm model configuration.
Validate the connection
Start with a short text request, then add streaming, images, tools, or structured output individually. Repeat these checks after changing the model or line, and inspect attribution and costs in usage. Model query example