API

How this differs from the desktop app

The desktop app runs the model locally on your own machine. Nothing leaves your computer, and your subscription covers unlimited use because you're supplying the compute yourself.

The API is the opposite arrangement: requests are served by our infrastructure, so they're billed by the token. The two are separate. Having a desktop subscription doesn't give you API credit, and using the API doesn't require a desktop subscription.

1. Create an API key

Keys are created and managed in the dashboard. Sign in, open the API keys tab, and create a key with a name that tells you where it's used.

The full key is displayed exactly once, at creation. We store a hash rather than the key itself, so if you lose it we genuinely cannot recover it for you, and you'll need to create a new one. After creation only the last four characters are shown.

2. Authenticate

Send your key in the Alden-Api-Key header on every request:

POST https://api.quantintelligence.co/v1/chat
Alden-Api-Key: YOUR_API_KEY
Content-Type: application/json

3. Make a request

A minimal example in Python. Read the key from the environment rather than writing it into the file:

import os
import requests

response = requests.post(
    "https://api.quantintelligence.co/v1/chat",
    headers={"Alden-Api-Key": os.environ['ALDEN_API_KEY']},
    json={
        "messages": [
            {"role": "user", "content": "Summarize this in one sentence."}
        ]
    },
    timeout=60,
)
response.raise_for_status()
print(response.json())

And the same request with curl:

curl https://api.quantintelligence.co/v1/chat \
  -H "Alden-Api-Key: $ALDEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"messages":[{"role":"user","content":"Hello"}]}'

4. Watch your usage

Every response reports how many tokens the request consumed, and the console aggregates this per key so you can see which integration is spending what.

Set a spend limit in the console's billing tab. Once you hit it, further requests are refused rather than continuing to bill, so a runaway loop in your own code can't quietly run up a large invoice.

Errors

Errors use standard HTTP status codes, with a JSON body explaining what went wrong and what to do about it. The ones worth handling explicitly:

• 401, the key is missing, malformed, or has been revoked.
• 402, you've reached your spend limit. Raise it in the console, or wait for the next period.
• 429, you're sending requests faster than your rate limit allows. Back off and retry.
• 5xx, a problem on our side. These are safe to retry with exponential backoff.

Rates and limits

Input tokens    $0.18 per 1,000,000
Output tokens   $0.43 per 1,000,000

You're billed on what you actually use, with no minimum. Output tokens cost more than input because generating them takes more compute than reading them, so prompts are cheap and long responses are what move the bill.

Your current rate limits are shown in the console against your account. Billing for API usage is separate from any desktop subscription and is processed by Paddle, our merchant of record.

Privacy

API requests are processed on our infrastructure, which is a meaningful difference from the desktop app's local-only model. What we retain, and for how long, is set out in Alden's privacy policy. If keeping data entirely on your own machine is the requirement, the desktop app is the right choice rather than the API.