Frequently asked questions
Practical answers about API compatibility, connected models, billing, logs, data processing, and support.
WLRouter is an OpenAI-compatible AI model gateway for accessing selected leading global and China-ready models through one API, with unified billing, logs, and usage tracking.
Yes. Use the OpenAI SDK with the WLRouter base URL, your WLRouter API key, and a connected model ID such as deepseek-v4-flash.
Only connected and verified models appear in Models, Playground, and /v1/models. Additional models are released after compatibility, billing, and reliability verification.
New accounts may receive a small trial credit for integration testing. Trial credit is consumed before prepaid balance and should be used with models that are online for the workspace.
WLRouter uses USD prepaid balance. Requests deduct usage cost from the available balance, and Billing shows top-up amount, service fee, credited balance, payment records, and receipt or invoice status when available.
Yes. You can top up by card through Stripe when checkout is available for your account and region. If checkout fails, contact support and we can help add balance manually.
Top-ups include a 6% service fee, with a minimum fee of $0.80. The fee and credited balance are shown before payment confirmation or support-managed balance changes.
Yes. OpenAI-compatible streaming is supported for chat completion models that can stream. Use stream: true in the request body and handle server-sent events in your client.
No. Failed requests are not billed. Successful requests are charged from verified usage when available, and each charged request is traceable in logs and billing records.
Console logs store request metadata for troubleshooting, billing, and abuse prevention. WLRouter never retains prompt or response content.
User-facing pages may show both the model company and the active supplier provider, such as Google with Google Cloud or Anthropic with AWS Bedrock. They do not show internal channels, upstream model IDs, base URLs, credential state, or fallback routing details.
Keep the request_id from the API response or error, then check Console Usage Logs or contact hi@wlrouter.com with the request ID and account email.