Home → FAQ
Fifteen questions people actually ask before and just after their first payment. Where the honest answer is “we do not offer that”, it says so.
APICLAN, at apiclan.us, is an OpenAI-compatible AI API gateway. One base URL gives access to Claude, GPT, Gemini, Grok and image models, billed per token in US dollars. It is a paid relay, not a model provider — requests are forwarded to the upstream providers and the models are their own.
Upstream capacity is purchased differently from a metered API account, and the saving is passed through as a published fraction of each model's official rate. The model, context window and speed are the provider's own — the price is what differs. Each model's page shows the official rate beside what you pay.
https://apiclan.us/v1 for OpenAI-lineage clients and SDKs. https://apiclan.us with no /v1 for Claude Code and the Anthropic SDK, because those clients append /v1/messages themselves. Getting it backwards produces a 404 at the edge, which means the request never reaches the gateway and never appears in your usage log.
That model is in a different group. A key belongs to exactly one group, so a key made for text models cannot call image models. The fix is a second key with the right group selected, not a permission change.
The trial credit is already in your account when you sign up — run real requests first. If the models do not behave the way you need, you have spent nothing.
No. Requests pass through the gateway to the model provider and the content is not written to our database — we keep only metadata for billing. The provider serving the model receives your prompt and applies their own policy.
Requests to that model fail until the provider recovers. There is no cross-provider fallback here. We do not publish uptime numbers we cannot back up; what we do promise is that support will tell you honestly what is broken and when we expect it back.
Top-ups are USDT on TRON (TRC20) only. Send the exact amount shown — the gateway matches your order by that amount. Unused balance is refundable within 7 days of a top-up, returned in USDT minus the network fee. Credits already spent are not refundable. Balance does not expire.
Send us the transaction hash and we will sort it out manually. The same applies if you paid after the payment timer ran out.
No. There is no subscription, no monthly minimum, no per-request token cap and no commitment to negotiate. A month where you build nothing costs nothing.
Prompt caching. Cache reads are billed at a fraction of the input rate and cache writes at more than it, so the same conversation costs very different amounts depending on whether context was reused or rebuilt. On our own traffic the median claude-sonnet-5 call costs 0.0087 credits while the mean is 0.0255 — a threefold spread on one model.
Keys are stored hashed and shown once when you create them. Anyone holding a key can spend your balance, so treat it like a password and revoke it in the console the moment you suspect it leaked.
Anything that accepts a custom OpenAI-compatible endpoint: Claude Code, Codex, Cursor, Cline, Roo Code, Cherry Studio, Chatbox, LobeChat and LM Studio, plus the official OpenAI SDKs for Python and JavaScript. Per-client settings are in the Quickstart.
No uptime SLA, and payment is USDT rather than card or invoice. If your organisation requires a signed vendor agreement or a data processing agreement before it can use a service, going direct to the provider is the right call — that trade-off is set out in full on our comparison page.
Email support@apiclan.us. A person reads it, usually within a few hours.
Prices and model availability change; the numbers quoted above are regenerated from live billing data on the price list, which is the authoritative version.
No subscription, no monthly minimum, no sales call. Top up with USDT and spend what you use — 1 USDT gives you 2 credits of API balance.
Read the 30-second quickstartPrices quoted on this page are regenerated automatically from live billing data. Third-party terms are quoted from that party's own published documentation.