Skip to content
Glacier AI Logo Glacier AI Logo Glacier AI Docs

Quick Start

This page takes the shortest path to a working model call. Once it works, read the individual feature pages in the User Guide as needed.

Prerequisites

Section titled “Prerequisites”
  • A Glacier AI account. Registration may ask you to verify your email address.
  • Available quota on the account. If a new account shows a zero balance, top up or redeem a code in the Wallet first.

Examples on this page use https://api.example.com as the site address — replace it with the address you actually use.

Step 1: Register and sign in

Section titled “Step 1: Register and sign in”
  1. Open the site home page and click Sign in in the top right, or go straight to /sign-in.
  2. If you do not have an account, click Sign up or go to /sign-up, then enter a username and password and complete email verification and the human check if required.
  3. After signing in you land on the console. The left navigation shows the modules available to your account.

See Sign Up & Sign In for details.

Step 2: Create an API key

Section titled “Step 2: Create an API key”
  1. Click API Keys in the left navigation, or go to /keys.
  2. Click Create API key and enter a name (name it after its purpose, for example “Production” or “Testing”).
  3. Set quota, expiry, model restrictions, group and IP allowlist as needed. Leave everything blank if you are unsure — blank means no restriction.
  4. After saving, open the list to view and copy the full key. It looks like sk-xxxxxxxx.

See API Keys for details.

Step 3: Send your first request

Section titled “Step 3: Send your first request”

The platform exposes an OpenAI-compatible API. Point the official SDK’s base URL at your site address plus /v1.

curl
curl https://api.example.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxxxxxxx" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello"}]
  }'
Python (OpenAI SDK)
from openai import OpenAI


client = OpenAI(
    api_key="sk-xxxxxxxx",
    base_url="https://api.example.com/v1",
)


response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

If you would rather not write code, click Playground in the left navigation (/playground) to pick a model and chat directly in the browser.

Step 4: Confirm usage and billing

Section titled “Step 4: Confirm usage and billing”
  1. Click Usage Logs in the left navigation, or go to /usage-logs/common.
  2. Find the request you just made and check the model, token counts, latency and quota charged.
  3. Click the record to expand the details and see the request ID, model mapping, billing breakdown and more.

See Logs & Async Tasks for details.

Common problems

Section titled “Common problems”
Symptom Where to look
401 Unauthorized Check the Authorization: Bearer <key> format, and whether the key is enabled and not expired.
403 Forbidden Check the key’s model restrictions, group permissions and IP allowlist.
Insufficient quota Check the balance or subscription quota in the Wallet, plus the quota cap set on the key itself.
Model not found Check in the Model Marketplace that the model is enabled and available to your current group.
No record in the logs Log writes lag by a few seconds, so wait and refresh. Not every failed response produces an error record.

Next steps

Section titled “Next steps”