Call the aijiau LLM API with an OpenAI-compatible client

The public English name is LLM API Gateway / Unified AI API. Chinese: AI+U / aijiau 模型 API. The console is api.aijiau.com. The www explainer is /gateway. This note is a getting-started path, not a second /jev column.

Start with one key

1. Sign up at https://api.aijiau.com/.
2. Create an sk-... key in the console.
3. Live catalog, balance, and available models are whatever the console shows that day. This page does not paste a stale model table.

Public CS is [email protected].

The OpenAI-compatible chat door

Chat models use:

POST https://api.aijiau.com/v1/chat/completions

The request shape is OpenAI Chat Completions: Authorization: Bearer sk-..., with model and messages in JSON. Most OpenAI SDKs or proxies can keep one key and point baseURL at https://api.aijiau.com/v1.

Use the model ids listed in the console. This page does not invent a catalog.

Jev does not use that door

Jev is TypeSafe’s System One decision model: state plus typed questions in; Choice / Score / Noul plus probabilities out. It does not write replies.

• Correct: POST https://api.aijiau.com/v1/systemone with model jev. That route is live.
• Wrong: POST /v1/chat/completions with model: jev. The live host returns 400.

The first-call walkthrough for the decision door is Land your first POST /v1/systemone. Pasteable shapes, the field map, and the Chinese self-test live on /jev and /jev/docs. Do not mix the two JSON shapes in one request.

What this page does not claim

• No unpublished QPS, concurrency, or SLA.
• No mashed “we are faster” table of vendor list prices versus aijiau customer prices.
• No tax quoting and no internal ops slang.

Next: open the console for a key; to land the decision layer, read Land your first POST /v1/systemone; or read /gateway for the product split.