Quickstart

The InferaStack Gateway is an OpenAI-compatible API. If your code already works with OpenAI, point it at our base URL and it works with InferaStack — no SDK to install, no request format to learn.

1. Get an API key

Keys are provisioned per organisation with a daily budget cap. Contact us to get set up — onboarding takes a day, and AWS Marketplace self-service subscription is coming soon.

2. Make your first request

curl https://api.inferastack.ai/v1/chat/completions \
  -H "Authorization: Bearer $INFERASTACK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "inferastack-chat-smart",
    "messages": [{"role": "user", "content": "Hello from Sydney"}]
  }'

3. Or use an OpenAI SDK

Set base_url to https://api.inferastack.ai/v1 and pass your InferaStack key:

from openai import OpenAI

client = OpenAI(
    base_url="https://api.inferastack.ai/v1",
    api_key=os.environ["INFERASTACK_API_KEY"],
)

response = client.chat.completions.create(
    model="inferastack-chat-fast",
    messages=[{"role": "user", "content": "Summarise this in one line: ..."}],
)
print(response.choices[0].message.content)

Model aliases

You call a stable alias; the gateway routes to the right backend. Swapping the underlying model is our config change, not your migration.

Alias	Use for	Routing
`inferastack-chat-fast`	High-volume, latency-sensitive tasks — drafts, classification, extraction	AWS Bedrock (Sydney gateway)
`inferastack-chat-smart`	Complex reasoning, longer generations	AWS Bedrock (Sydney gateway)
`inferastack-chat-or-claude`	Frontier-model quality via the catalog path	OpenRouter (routes via US)
`inferastack-chat-or-llama`	Open-weight workloads via the catalog path	OpenRouter (routes via US)

Data residency: inferastack-*-fast/smart aliases are invoked from our Sydney (ap-southeast-2) gateway via AWS Bedrock. *-or-* aliases route through OpenRouter in the US — choose per request based on your data requirements. Sovereign private deployments pin all traffic to your environment.

Costs and budgets

Every response is metered: the gateway computes the exact cost of each request and records it against your key, and daily budget caps stop runaway spend before it happens. Ask us for cost reporting when you onboard.

Errors

Chat endpoints return OpenAI-shaped errors, so existing retry and error-handling logic keeps working:

{
  "error": {
    "message": "Daily budget exceeded for this key.",
    "type": "insufficient_quota",
    "code": "budget_exceeded"
  }
}

Deployment options

The hosted gateway at api.inferastack.aiis the fastest way to start. The same gateway deploys privately — in your VPC, on-premise, or in NEXTDC Tier IV facilities — when data can't leave your environment. As an AWS Partner, we're bringing the gateway to AWS Marketplace so you can subscribe and pay through your existing AWS bill. Talk to us about which fits.