Chinese AI model routes

Choose the model for the workload, not the logo.

Evaluate Chinese AI model routes through one OpenAI-compatible API, one USD billing surface, and a dated request-level ledger.

Create a test key Check dated pricing

Start with a request you can verify.

Keep the model ID, status, usage fields, and effective group beside the test result.

One request before production traffic

FITCHECK

Start with the job to be done.

Define context length, output shape, tool use, latency tolerance, and acceptance criteria before selecting a route.

TESTCHECK

Send one representative payload.

A small hello-world call checks credentials. A representative payload checks whether the model and route fit your application.

CONTROLCHECK

Keep the model ID explicit.

Record the tested model ID with the result so a later route change cannot silently alter your acceptance evidence.

curlOpenAI-compatible chat request

curl https://aiwave.live/v1/chat/completions \
  -H "Authorization: Bearer $AIWAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4-pro","messages":[{"role":"user","content":"Reply with one sentence."}]}'

Questions to settle before you send traffic.

The answers keep account, route, and billing boundaries explicit.

How do I compare Chinese AI models?

Use the same representative payload and acceptance checks for each route. Compare quality, context fit, latency, failure behavior, and dated cost.

Can I keep my existing OpenAI SDK?

For supported chat requests, set the AIWave base URL and key, then choose an AIWave model ID.

Does one listed route guarantee availability?

No. Catalog visibility is not a guarantee that every request will complete. Keep retry limits and fallback decisions in your application.