Getting started
QDivZero is a managed inference platform. You choose a model. We provision a compatible GPU, prepare the infrastructure, and publish an endpoint that works with the OpenAI API. This guide takes you from account creation to your first successful request. Deployment time depends on the model and the available capacity.
Before you start
1. Create your account
Open the registration page and create your account. QDivZero creates a default workspace, assigns an account ID, and gives the first member the owner role. You can invite more people later from the account settings.
2. Add credit
Compute is billed for each hour that a GPU remains active. Go to Billing and top up, choose an amount in euros, and complete the secure checkout. The balance history records every credit purchase, charge, and refund so you can see where your credit goes.
Start with a small balance
3. Create an API key
API keys authenticate every request sent to QDivZero. Open API keys and create a key. The complete token appears only once. Store it in your secret manager before you leave the page.
1# Store the key in the current shell
2export QDIV0_API_KEY="qdiv0_sk_..."4. Deploy your first instance
An instance is a deployed model with a public serving name. Open Create an instance and choose a compatible Hugging Face repository. QDivZero checks its memory requirements, selects a suitable GPU, and makes the model available through the OpenAI API.
- Use Manual mode when you want to choose the GPU yourself. Use Smart mode when you want the scheduler to select it.
- Choose a short serving name that is easy to recognize. For example
qwen35-demo. - If the workload supports chat completions, you can attach a firewall to inspect requests before they reach the model.
5. Send your first request
When the instance reaches running status, it is ready to receive requests through the OpenAI API. Use https://api.qdiv0.com/v1 as the base URL. The model name is the serving name you chose during deployment.
1from openai import OpenAI
2
3client = OpenAI(
4 base_url="https://api.qdiv0.com/v1",
5 api_key="your-api-key",
6)
7
8response = client.chat.completions.create(
9 model="qwen35-demo",
10 messages=[{"role": "user", "content": "Hello, world!"}],
11)
12print(response.choices[0].message.content)To generate embeddings, send requests to /v1/embeddings with an instance created for embeddings.
