VensixOne API key for every AI model.
Gemini Flash, GPT-4o mini, DeepSeek V3, Kimi K2, Claude Haiku, GPT-4o, and Claude Sonnet 4.5 — reachable through a single OpenAI-compatible endpoint, billed per token in Indonesian Rupiah.
What the Vensix AI API is
Vensix is an API gateway that speaks the OpenAI chat completions format. One key — starting with sk-vsx- — unlocks every model in the catalog. Point your tool at our base URL and keep the integration you already have: it works in Cursor, Cline, Claude Code (via base URL override), the official OpenAI Python and Node.js SDKs, LangChain, or any plain HTTP client.
It is honest by design. When an upstream model fails, the error is passed through exactly as the provider returned it — Vensix never fabricates a completion to hide a failure. Your balance is only charged for successful responses, metered from the actual token counts and the actual upstream cost of that call. The same record is what you see in your dashboard history — no sampling, no estimates.
Billing is prepaid and local: top up your Rupiah balance via QRIS — GoPay, OVO, ShopeePay, DANA, or any QRIS-enabled wallet — then spend it per token at the live prices shown below. The same balance also pays for Cloud Hosting and every Vensix product that launches next.
Common use cases
Coding assistants
Power Cursor, Cline, or Claude Code with the model that fits the task — from Gemini Flash for quick edits to Claude Sonnet 4.5 for deep refactors.
Chatbots & support
Build customer-facing chat in Indonesian or English, with per-token costs you can predict before you ship.
Content & research
Draft, summarize, and translate at scale with GPT-4o, DeepSeek V3, or Kimi K2 — swap models by changing one string.
Automation & agents
Chain models inside your own pipelines over a plain HTTP API — no SDK required, no vendor lock-in on the client side.
The model catalog
Seven models behind one key, fetched live from the same catalog that bills you.
How it works
Create your key
Sign in and generate an sk-vsx- API key from your dashboard — one key works for every model in the catalog.
Top up via QRIS
Add Rupiah balance with any QRIS e-wallet. Your balance updates automatically once the payment is confirmed.
Point your tool to our base URL
Set the base URL and key in Cursor, Cline, the OpenAI SDK, or your own code — requests are billed per token at live prices.
Live model pricing
Per 1 million tokens — input & output priced separately, fetched live from the same catalog that bills you.
Prices follow the live USD–IDR exchange rate and current upstream costs. Full pricing, including hosting, is on the home page.
AI API questions
Which tools work with the Vensix AI API?
Anything that speaks the OpenAI chat completions format: Cursor, Cline, Claude Code (via base URL override), the official OpenAI Python and Node.js SDKs, LangChain, or plain curl. Set the base URL to our endpoint and use your sk-vsx- key — no other changes.
What happens if a model fails mid-request?
The upstream error is passed through to you exactly as the provider returned it — never a fabricated answer. A failed request does not charge your balance; only a request that successfully returns a response is billed.
How am I billed?
Prepaid, per token. Top up your Rupiah balance via QRIS, then each request is charged from the actual token counts of that call at the live per-model prices shown above. There is no subscription and no automatic monthly billing.
Why can model prices change over time?
Prices follow the live USD–IDR exchange rate and the current upstream provider costs, converted at the moment you use them. The prices on this page are fetched live from the same catalog the billing uses, so what you see is what you pay.