One key. Every model.Nothing waits on a single provider.
sjetZ sits in front of every AI provider you connect and gives you one endpoint, one key, and one balance — metered to the token, routed around outages automatically.
Everything a gateway should do
sjetZ is built on New API, so the plumbing a serious integration needs is already there — not bolted on later.
Unified endpoint
One base URL and one key reach every OpenAI-compatible provider you connect, with no per-provider client changes.
Usage-based billing
Every call is metered to the token, with top-ups and redemption codes for teams.
Load-balanced routing
Traffic spreads across channels per model and fails over the moment one stalls.
Chat, image & video tasks
Longer-running generation jobs run as trackable async tasks.
Full request logs
Every call is recorded with cost, latency, and model — down to the key.
Team accounts
Invite teammates and let everyone draw from one shared balance.
One request, routed until it lands
Every call takes the same three hops — your app, the gateway, the provider — with health checks and failover built into the middle one.
Your app
Calls the OpenAI-, Anthropic-, or Gemini-shaped endpoint you already integrate — no new SDK to learn.
sjetZ gateway
Meters the call, picks a healthy channel for that model, and load-balances across every channel behind it.
42+ providers
The request lands on the provider and model you asked for — and fails over instantly if that channel stalls.
Same request format you already know
Point your existing OpenAI, Anthropic, or Gemini client at sjetZ's endpoint. No new SDK, no rewritten request bodies.
# same OpenAI-compatible request shape, any model behind it
curl https://api.sjetz.com/v1/chat/completions \
-H "Authorization: Bearer $SJETZ_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-opus-5-5",
"messages": [{"role": "user", "content": "Summarize this outage report."}]
}'from openai import OpenAI
client = OpenAI(
api_key="sk-sjetz-••••••••••••",
base_url="https://api.sjetz.com/v1",
)
response = client.chat.completions.create(
model="claude-opus-5-5",
messages=[{"role": "user", "content": "Summarize this outage report."}],
)
print(response.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
apiKey: "sk-sjetz-••••••••••••",
baseURL: "https://api.sjetz.com/v1",
});
const response = await client.chat.completions.create({
model: "claude-opus-5-5",
messages: [{ role: "user", content: "Summarize this outage report." }],
});
console.log(response.choices[0].message.content);
See exactly what's happening, request by request
Every call is metered, logged, and traceable — down to the model, the cost, and the millisecond.
One gateway, three ways to use it
Ship a side project against several models without juggling separate accounts and bills for each provider.
Resell metered model access to your own users under one brand, with redemption codes and top-ups handling billing.
Give your company one gateway and one audit trail instead of scattered provider keys.
Open by design, secured by default
sjetZ runs on New API's open-source core, so nothing about how it works is hidden from you.
Open-source core
Built on New API — inspect it, don't just trust it.
TLS everywhere
Every connection is encrypted in transit, end to end.
EU-hosted
Infrastructure runs on EU-based servers.
You own your keys
Rotate or revoke any API key instantly from your dashboard.
Simple, usage-based pricing
Top up credit once and spend it across every model sjetZ routes to. Exact per-model rates are set in your dashboard and shown before you spend anything.
- A separate account, key, and invoice per provider
- One outage stalls the whole app, with no failover
- A leaked key means finding and rotating it across N dashboards
- One key and one balance for every provider you connect
- Automatic failover to the next healthy channel
- Rotate or revoke a key instantly from one dashboard
No plans to pick, no seats to buy
You're billed for tokens sent and received, at the rate shown for that model — nothing else.
