Help center

Sign up, copy the default API key from your dashboard, paste the cURL on the Quickstart page into your terminal. Should work in under 5 minutes.

If unsure: qwen3-max for code, qwen3-max for SEA-language chat, gpt-4o-mini for high-volume simple tasks. Or browse models by use case.

Yes — $1 trial credit on signup, no card required. Enough for ~500 chat turns on gpt-4o-mini.

Pay-as-you-go and Enterprise plans get monthly Singapore billing records with custom headers and tax IDs where applicable.

Visa / Mastercard / Amex via Stripe. GrabPay, PromptPay, TrueMoney, GoPay, and GCash through local wallet payment links.

Unused topped-up balance refundable in full within 30 days. Consumed tokens not refundable.

Dashboard → Keys → Rotate. A new key is issued immediately; the old key is revoked immediately and returns 401 on next use.

Yes — set a monthly budget per key. When the cap is hit, that key returns 402 until next month or you raise the cap.

Per-account default 60 RPM. Pay-as-you-go: 300 RPM. Enterprise: custom limits. Email support if you need more.

SeaLink manages provider connections internally, with automatic fallback if a provider degrades. You use SeaLink's keys, not your own provider account — so your personal provider account status doesn't affect SeaLink service. Current availability is always shown on /status.

No. We log only metadata: model, token counts, latency, status. Request/response bodies are never persisted.

Singapore by default. Enterprise clients can discuss dedicated residency requirements during onboarding.

Yes — SeaLink is OpenAI-compatible. See the Cursor integration guide; we provide a one-click config download.

Yes, via two env vars. We offer an Anthropic-compatible endpoint that works directly with Claude Code.

All supported with OpenAI-standard syntax. See /docs/function-calling, /docs/vision, /docs/caching.