Questionsalotof peoplewantanswered.

Yes. Every successful top-up produces a receipt and Thai tax invoice in THB from I C Develop Co., Ltd., a company registered in Thailand — an abbreviated tax invoice by default, downloadable from your top-up history in the member portal. That is the reason this gateway exists: paying OpenAI, Anthropic or Google directly gets you a foreign receipt that your accounting team cannot file the same way.

For a full tax invoice, choose the full tax invoice option and enter your tax ID and address when you top up. Top-up amounts exclude 7% VAT, which is added at payment and shown separately on the document.

Not for what you spend here. Buying from IC Lueur Ai is a domestic purchase from a Thai company, so it is claimable at 100% as an ordinary expense with no foreign-service VAT self-assessment (ภ.พ.36) and no withholding remittance for a foreign payee (ภ.ง.ด.54). Paying a foreign AI provider directly is what creates those filings. This is a description of how the purchase is structured, not tax advice — your accountant signs the return, not us.

PromptPay from a Thai bank account, minimum ฿300 per top-up before 7% VAT. No foreign credit card is needed and there is no FX spread applied at top-up. Credit is prepaid and does not expire on a timer.

No. There is one prepaid balance and you are billed from measured usage — input, output and cached tokens are priced separately because the upstream providers price them separately. There are no tiers and no per-seat charge.

Nothing you already spent changes. The rate in force when a request ran is stored with that request, so a later price change never re-prices past usage. The billing ledger is append-only — corrections are posted as new compensating entries, never by editing history.

In Thailand. Your account, your credit balance, the append-only billing ledger, your usage records and your tax invoices are all held in the AWS Asia Pacific (Thailand) region, ap-southeast-7, along with the gateway and control plane themselves. An encrypted nightly backup is also kept outside AWS on Cloudflare R2, and system logs are sent to our company monitoring system.

Yes, in transit and at rest. Connections from your client to the gateway, and from our services to the database, cache and file storage, use TLS. Traffic between our own services stays on the private network inside AWS and is not TLS-encrypted. The PostgreSQL database, the cache and the storage holding invoices and reports are encrypted at rest, and API keys are kept only as a hash. The security page lists each measure.

The database keeps running. It has a synchronous standby in a second AWS availability zone in Thailand that takes over if the first zone fails, and it can be restored to any point in the last 7 days. Some other components still run as a single copy, so we do not offer an uptime guarantee.

No. AWS holds ISO 27001, SOC 2 and other certifications for the infrastructure we run on, but IC Lueur Ai itself is not certified to any of them, and our own measures have not yet been assessed by an outside security firm. The security page describes what we have in place instead.

Yes, through your organization’s routing policy, which our team sets for you on request. A new API key takes the organization’s policy at the time it is created, and the policy is enforced on every call made with that key. There are three:

• strict — Bangkok only, never another region. No model is served from Bangkok yet, so requests under strict are currently refused with a 400.
• regional — the default. The nearest region is tried first — Bangkok, then Singapore, then Tokyo — and a model not served in any of them goes to the provider’s global region, which can be outside Asia, for example the United States.
• available — only for organizations that ask for it. It reaches everything regional does plus providers that process in mainland China, and image and video models served outside Bangkok, Singapore and Tokyo require it.

Changing the organization’s policy does not change keys created before the change. See the docs for more detail.

No. The gateway records token counts, cost, model, provider, region and timestamps — the fields needed to bill you. Prompt and completion bodies are never written to a store or included in any payload that leaves the process, and we do not train anything on your traffic. The upstream provider you route to has its own terms, which we do not override.

No, and that is deliberate. A key is shown once when you create it and only a hash of it is stored, so we genuinely cannot display it again. Create a replacement and revoke the old one. The member portal also has two-factor authentication with recovery codes.

The base URL and the API key. The gateway speaks the OpenAI Chat Completions API, so an existing openai-python, openai-node or LangChain client — anything that lets you set a base URL — keeps working. Errors use the same shape as OpenAI’s, {"error": {"message": "..."}}, but the type and code values do not always match OpenAI’s — if your code branches on them, use the HTTP status and the message text instead.

Yes. Set stream: true and read the response as server-sent events, in the same chunked format OpenAI uses. Note that tokens generated before you disconnect are still billed — an interrupted request is metered like a completed one.

The live catalogue on the pricing page is the answer, and GET /api/models returns the same thing as JSON with no authentication.

It spans Anthropic Claude, OpenAI GPT, Google Gemini, Meta Llama, Mistral, DeepSeek and others, across AWS Bedrock, Google Vertex AI, Azure AI Foundry and Fireworks.

We do not publish a fixed count here because the catalogue changes.

Yes. An organization can have several people in it with a role each — owner, admin or member — so nobody has to share a password to see the balance or manage keys.

No. We charge provider list cost plus a margin, so per token we are usually more expensive than a direct contract. If you can hold your own agreement with AWS, Google or Microsoft and handle the foreign-service tax paperwork yourself, that route is usually cheaper. What the margin buys is the Thai tax invoice, PromptPay settlement, a routing policy for your organisation and support during Thai business hours. If price per token is the only thing that matters to you, we are not the right gateway.

Yes. Every API key has a requests-per-minute limit — one you set on the key in the member portal, or the gateway default if you leave it unset. A request over the limit gets a 429 with a Retry-After header; it is not charged and not counted as usage. Free-model promotions add their own per-key limit on top, usually tighter — the docs explain how to tell the 429s apart.

On the status page. It shows whether each region is up, degraded or down, recalculated roughly every 5 minutes.

Time-to-first-token per model comes from real requests sent through our gateway and is re-measured roughly every hour. A model that has not been measured yet shows "—" instead of a number.

Only the latest reading is shown: there is no uptime history or uptime percentage yet, and tokens-per-second is not published.