The AI Gateway for Developers

Build with multiple AI models through one API, with prepaid Coin, clear usage, and control over every request.

Create account Read the docs

Published now

What's in Chanx AI?

Everything you need to connect AI models to your product.

smart-code

CapabilityAvailabilityBudget

SmartRoute

Pin an exact model or choose a policy alias. Inspect the reason behind each routing decision.

Workspace

Project

API key

SmartBudget

Set spend limits for workspaces, projects and API keys before the next request starts.

…

Published models

Model Lab

Compare model capabilities and your workspace's real performance, with every source clearly labeled.

True streaming

Receive output as it arrives. Follow each request from the first chunk through final settlement.

ContextGuard

Inspect repeated context and approve conservative optimizations before changing a request.

Request Explorer

See tokens, routing, latency and exact charges together in a single request view.

Versioned pricing

Every charge uses the request's pricing version, with input, cache and reasoning accounted for separately.

Scoped API keys

Create project keys with model restrictions, expiration, budgets and immediate revocation.

Prepaid Coin

Choose a package, pay with QRIS and track the verified credit in your wallet ledger.

Your tools. Your models. One connection.

Connect your client to Chanx AI. Choose from the published catalog and keep your access, budget, and usage in one place.

Your app
OpenAI SDK
Claude Code

Chanx AI.

One endpoint. One balance.

...

Your models. One connection.

Published model families connected through Chanx AI. Explore the live catalog before choosing your route.

Explore models

Loading published model families

One request. Visible economics.

Request lifecycle

POST /v1/chat/completions

How billing works

Process overview. Your real requests and charges appear in Request Explorer.

Service
Published models
Top-up from

Understand the range

Model, context, reasoning level and output limits produce an estimate. It is never a guaranteed final charge.

estimated
reserved
actual
released

Keep your client. Change the endpoint.

Use a project key with Chat Completions or the supported Responses text API. Stream output and follow the request ID into the console.

Your endpoint: http://localhost:8080/v1

curl http://localhost:8080/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_PUBLISHED_MODEL",
    "messages": [{"role":"user","content":"Hello"}],
    "max_tokens": 128
  }'

No model is published yet, so the model is a placeholder. Create and store your key securely. Never embed it in browser code.

Get started with curl

Start with a clear balance.

Packages are read from the billing catalog. Pay through QRIS and receive Coin once the payment is verified.

From invoice to Coin

Choose a package
Check the exact amount and Coin grant first.
Pay the exact amount
The QRIS asks for the package price plus a small unique code.
Payment is verified
An administrator confirms it. A scan or redirect never counts.
Coin is credited
The wallet ledger records the credit with its invoice reference.

Pricing and billing details

Know what is available.

Published models and service status come straight from the gateway.

Published models

Service status

View service status

Treat keys and Coin as critical infrastructure.

Private by default
Prompts and responses are not stored by default. Request Explorer shows operational and billing metadata.
Scoped access
Project keys, workspace roles, model restrictions, expiration and revocation keep access explicit.
Auditable money movements
Reservations, settlements, releases and top-ups use an immutable ledger. Browser redirects never confirm payment.

Put your AI requests in view.

Create a workspace, check the catalog, and configure your project.

Create account

The AI Gateway for DevelopersTThhee AAII GGaatteewwaayy ffoorr DDeevveellooppeerrss