Skip to content
1link AI

One API key for every AI model

Claude, GPT, Gemini, DeepSeek, and many more models through one endpoint and one Rupiah balance. Pay only for the tokens you use.

Rupiah balance, top up through an administratorNo subscription

Get started freeView documentation

Model providers available through Satulink AI

  • OpenAI
  • Anthropic
  • Gemini
  • DeepSeek
  • Meta
  • Qwen
  • Kimi
  • xAI

Three steps to your first request. Create a key, choose a model, then track costs — all in one dashboard.

  1. Create an API key

    Create one key on the API Keys page. Add an optional name and restrict models, RPM, concurrency, and IP addresses when needed.

  2. Choose a model, keep your code

    Change one model line on the OpenAI-compatible endpoint. Python, Node.js, and cURL all use the same format.

  3. Track costs & requests

    Every request records its tokens, latency, status, and cost — deducted from your Rupiah balance.

Catalog & rates

Models & pricing in Rupiah.

Rates per million tokens, loaded from the public catalog with this page. The actual cost of every request appears on the dashboard's Usage page.

The catalog could not be loaded

The model list and pricing will appear here when the gateway responds. In the meantime, the documentation explains how to call a model.

02 · Endpoint & catalog

One endpoint. Change models, keep your code.

Choose a public model below. The demo will type a request, wait, then stream a deterministic response with pricing and statistics.

Learn about the chat endpoint

The model catalog is unavailable; the example below uses a placeholder ID.

Your selected modelResponse

Summarize the main AI trends of 2026 in 3 short points.

Waiting for the model…

Tokens
0
Latency
—
Estimate
IDR 0

Sample request: 1,000 input tokens + 500 output tokens, priced at the model's rate. Every use is recorded on the Usage page.

Illustration — this page does not send a request to the gateway.

03 · Access control

One key, with limits you define.

Add an optional name, allow all or selected models, then set RPM, concurrency, IP addresses, and expiration. The full secret appears only once after creation.

Manage keys in the dashboard

04 · Costs & requests

Trace every Rupiah to the right request.

Track cost, request count, tokens, usage by model, and each request's status, latency, and cost. Values in the frame below are static illustrations, not live data.

Open the Usage page

FAQ

Frequently asked questions

Still have questions? The documentation covers the technical details of every endpoint, from authentication to billing.

No. You only need one Satulink AI account and one API key. The same key can call models from OpenAI, Google AI (Gemini), Anthropic, and other providers in the catalog — without signing up with each provider.

Coordinate a Rupiah top-up with an administrator. Once confirmed, the balance is ready to use. There is no subscription or monthly fee.

Each model has input and output rates per million tokens. The charge is based on the tokens actually used, deducted from your balance, and recorded per request on the Usage page.

Yes. The format follows the OpenAI API: change the base URL to https://api-dev.satulink.ai/v1 and use your Satulink API key. The openai SDK for Python and Node.js, cURL, and other tools using the same format work directly.

Requests are rejected until you top up. Your API keys and usage history remain intact — once an administrator confirms the top-up, the same keys work again immediately.

One API key. Every model.

Create an account, coordinate a Rupiah top-up with an administrator, then send your first request with one API key.

Get started freeView documentation