Quickstart
Sign in to the Console and send your first Gateway request.
This guide walks through the complete path from signing in to making an AI Gateway request.
1. Sign In to the Console
Open the Console. Existing users can sign in with their company email address and password. New accounts must verify their email address with a code that remains valid for five minutes.
2. Issue a Gateway Key
Create a Gateway Key on the Gateway Keys page and configure:
- Allowed models: Use
*to allow every available model, or specify individual model aliases. - Rate and concurrency limits: Set requests per minute (RPM), tokens per minute (TPM), and maximum concurrency.
The complete key is shown only once when it is created. Store it securely.
3. Send a Request
Call AI Gateway through its OpenAI-compatible API:
curl https://api.centidor.ai/v1/chat/completions \
-H "Authorization: Bearer $CENTIDOR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-opus-4-8",
"messages": [{"role": "user", "content": "Hello"}]
}'4. Review Usage
Use Usage to audit the lifecycle, token consumption, and charge for each request. Wallet shows the available balance and top-up orders.
Next step
Need to isolate keys and usage by workload? Create a separate Project for each workload on the Projects page.