MODELS
Unlimited Basic subscription: unlimited Qwen 3.8 27B for €14.95/month. GLM + DeepSeek with 5,000 shared AI Credits. Fair use applies.

Documentation

Guides for deploying models, connecting the OpenAI-compatible API, and configuring routing, protection, and retrieval.

Documentation

Run AI models without assembling the infrastructure around them.

QDivZero is an AI infrastructure platform for taking compatible models from Hugging Face and other supported sources into production. It finds GPU capacity, prepares the runtime and storage, operates the deployment and exposes it through an OpenAI-compatible API. The same platform also provides ready-to-use models, LLM Model Routing, LLM Firewall and Retrieval.

Deploy your first modelOpen the API reference
platform.qdiv0.com
Deployments
Models you have launched in your account.
How deployments work
Deployments
9
In your account
Active deployments
1
Ready to receive traffic
Estimated total per second
€0.00016389
Active instances only
DeploymentsModel catalog
Search by name, ID, model or status
NameTypeStatus
Qwen3.8-27B-NVFP4 instanceChatActive
Qwen3.8-27B-NVFP4 instance
e3d00c38-8ec9-4946-8022-36f935ad57a0
Details
ID
e3d00c38-8ec9-4946-8022-36f935ad57a0
Infrastructure
Intelligent
Region: North America, Europe

What changes with QDivZero

The difference is not another model endpoint. It is having the execution and operational layers behind that endpoint under one contract.

01

GPU provider or marketplace

A machine or raw accelerator capacity.

Your team validates the model, builds the runtime, manages storage, exposes serving and operates every deployment.

02

Managed model API

A ready endpoint from the provider's catalog.

Your model choice, pricing and behavior stay tied to that provider; changing execution can require another integration.

03

QDivZero

Model choice, multi-provider capacity, runtime, storage and a stable OpenAI-compatible endpoint.

You choose the model, capacity constraints and policies. QDivZero operates the shared platform behind them.

The QDivZero platform

Start with the layer you need. Every product uses the same account, API keys and operational model.

01DeploymentRun compatible models with capacity selected across providers, persistent storage and active-capacity pricing.02ModelsCall selected ready-to-use models with published usage pricing and the same inference API.03Inference APIUse one OpenAI-compatible contract across deployments, models and routing endpoints.04LLM Model RoutingRoute by intent or priority and fail over between healthy destinations behind one serving name.05LLM FirewallEvaluate supported chat requests with reusable policies before they reach the model.06RetrievalTurn QDivZero embeddings into semantic search, recommendations and context for RAG.

Built for the whole AI workflow

For companies

Consolidate model execution, usage, policies and provider decisions in one place, with clear capacity and metered pricing.

For AI and product teams

Test models and architectures without rebuilding deployment, routing and protection for every experiment.

For developers

Keep familiar OpenAI SDKs and request formats while QDivZero manages the infrastructure and control plane behind them.

Choose where to start

I want to make the first API callCreate a key, deploy a model and connect the OpenAI SDK.I want to deploy a modelUnderstand capacity selection, lifecycle, storage and pricing.I already know what I needGo directly to endpoints, methods and request examples.