Earn from your idle GPU

Your GPU works while you sleep.

GPURenta connects your idle GPU to AI inference jobs. No Docker. No WSL2. No SSH. Just install and earn. 10% commission. Stripe payouts to your card.

Download for Windows Download for macOS
Windows 10+ macOS 12+ Linux x86_64 Apple Silicon
0
GPUs online
0
Earned this month
10%
Commission

Works with

Ollama GGUF PyTorch CUDA TensorRT ONNX OpenAI API

Income calculator

How much will you earn?

12h
6₽
85%
Gross revenue ₽0
GPURenta fee (10%) -₽0
Electricity -₽0
After fee ₽0
Your net earnings / month ₽0

Based on average AI inference demand. Actual earnings vary by workload availability, GPU demand, and uptime.

Getting started

Three steps to start earning.

Install the agent

Download GPURenta. It auto-detects your GPU, installs Ollama, and connects to the network. No config. No terminal.

Let it run

The agent runs silently in your system tray. When your GPU is idle, it accepts AI inference jobs. When you game or work — it pauses automatically.

Get paid monthly

Earnings accumulate in your dashboard. Withdraw anytime via Stripe — money lands on your card. No gift cards. No crypto.

Comparison

GPURenta vs other platforms.

GPURenta Griddly Salad GamerHash Owlrun
Commission 10% ~20% 20–25% Variable 4–9%
Payout method Stripe → your card Gift cards Gift cards, PayPal GHX token, USDC Bitcoin Lightning
Platforms Win + Mac + Linux Win (Mac soon) Windows only Windows only Win + Mac + Linux
Architecture Native binary WSL2 + Docker WSL2 + Docker Native binary Native binary
SSH access Never No No No No
Auto-pause on game Yes Yes No Yes Yes
Min GPU RTX 3060 (12GB) RTX 3060 Any NVIDIA RTX 3060 Any CUDA

Security & trust

Your PC stays yours.

No Docker. No WSL2.

Workloads run as native binaries managed by the agent — no virtual machines, no container escapes, no kernel exploits.

No SSH. No shell access.

Developers send jobs through our API. They never get a terminal, never touch your files, never see your system.

Pre-audited workloads.

Every workload image is reviewed before it reaches your GPU. Custom code runs in a sandbox with no filesystem access.

For developers

Inference API at 70% lower cost.

Use the OpenAI-compatible API to run inference on distributed consumer GPUs. Pay per second. No minimum. No contract.

  • → Drop-in replacement for OpenAI — change one line of code
  • → Models: Llama 3, Mistral, Qwen, DeepSeek, SDXL, Whisper
  • → From 8₽/GPU/hour vs 35₽+ on AWS/GCP
Get started
# OpenAI-compatible — swap one line

from openai import OpenAI

client = OpenAI(
  base_url="https://api.gpurenta.com/v1",
  api_key="gpr_..."
)

response = client.chat.completions.create(
  model="llama-3.2-8b",
  messages=[
    {"role": "user", "content": "Hello!"}
  ]
)

# → runs on distributed GPU in ~2 seconds
# → costs ~0.03₽ per request

Early users

Trusted by GPU owners and developers.

"I leave GPURenta running on my gaming PC overnight. Earns me about $80/month — covers my Steam purchases and then some. Setup took two minutes."

MK
Mike K.
RTX 4090 owner

"We cut our inference costs by 60% after switching from RunPod. Same models, same latency, but running on distributed GPUs at a fraction of the price."

AL
Alex L.
ML Engineer

FAQ

Frequently asked questions.

Earnings depend on your GPU, uptime, and current demand. An RTX 4090 running 12 hours/day earns roughly $80–120/month. An RTX 3060 earns $25–40/month. Use the calculator above for a personalized estimate.

No. The agent monitors GPU usage and keyboard/mouse activity. When you launch a game or start working, it pauses all inference jobs instantly. Your GPU is always yours first.

Via Stripe Connect. Link your card or bank account once. Earnings are paid out monthly, with no minimum threshold. You receive real RUB — not gift cards, not crypto, not in-app tokens.

The agent runs pre-audited binaries only — no containers, no SSH, no shell access. Your personal files are never touched, never transferred, never accessible to anyone.

NVIDIA RTX 3060 (12GB) and above. Apple Silicon M1 and above (via Metal). AMD ROCm support coming soon. Minimum 8GB VRAM recommended.

Yes. GPURenta works on gaming laptops with NVIDIA Optimus. The agent uses the discrete GPU and only runs when plugged into AC power. Battery mode is paused automatically.