Introduction to GPUniq

GPU Meta-Cloud platform — rent GPUs, deploy containers, manage persistent storage, and access LLM models through a unified API and Python SDK.

Overview

GPUniq is a GPU Meta-Cloud platform that aggregates compute capacity from multiple providers worldwide. Access thousands of GPUs through three deployment modes, manage persistent storage volumes, and use LLM models — all through a single API and Python SDK.

Save up to 70% compared to traditional cloud providers. No upfront commitments, pay only for what you use.

GPU Compute

GPUniq offers three ways to rent GPU compute, each suited for different use cases.

GPU Marketplace

Browse 5000+ GPU servers. Pick the exact machine, compare specs, and rent with full control.

GPU Dex-Cloud

Deploy by GPU type. Pick a model, click deploy — the platform finds the best server automatically.

GPU Burst

Scale to 100 GPUs with automatic fallback. Best for distributed training and batch inference.

Platform Features

Persistent Volumes

Storage that survives instance restarts. Upload data, attach to any instance.

LLM API

Claude, GPT-5, Gemini, Grok, and open-source models — one API key. OpenAI-compatible endpoint works with Claude Code, Cursor, and LiteLLM.

Python SDK

Full-featured Python SDK. pip install GPUniq and control everything programmatically.

CLI (gg)

Checkpoint and replay commands on GPU instances. Survive restarts without losing progress.

Quick Start

Create Account

Get API Key

Go to LLM API Keys in your dashboard and create a key. It starts with gpuniq_.

Install SDK

pip install GPUniq

Deploy a GPU

from gpuniq import GPUniq

client = GPUniq(api_key="gpuniq_your_key")

# Deploy an RTX 4090 instance
deploy = client.gpu_cloud.deploy(
    gpu_name="RTX_4090",
    docker_image="pytorch/pytorch:latest",
)

Authentication

GPUniq supports two authentication methods:

API Key (recommended for SDK and automation): Pass via X-API-Key header
JWT Token (web dashboard): Pass via Authorization: Bearer header

from gpuniq import GPUniq
client = GPUniq(api_key="gpuniq_your_key")

Rate limit: 120 requests/minute per API key. The SDK retries automatically on rate limit.

Next Steps

Quickstart Guide

Step-by-step account setup and first GPU rental.

Python SDK

Install the SDK and start building.

Was this page helpful?

Last updated today

Built with Documentation.AI