Rent an NVIDIA A100 80G by the minute

The NVIDIA A100 80G is a data-center card with 80 GB of memory. It is for jobs that need more memory on one GPU than consumer cards offer. On CloudGPU you rent it by the minute, not by the hour, and destroy it when the job is done.

See the current A100 80G rate · Deploy an A100 80G

The card

A100 80G
Architecture Ampere
VRAM 80 GB
Memory bandwidth 1935 GB/s
FP16 312 TFLOPS
CUDA compute capability 8.0
Billing Per minute, 1-minute minimum
Rate On cloudgpu.app/pricing

What it is for

Which model fits depends on its size, quantization and context length; see how much VRAM an LLM needs.

Related guides: Llama 3.3 70B, vLLM on a rented GPU and Rent a GPU for LoRA fine-tuning.

If your workload fits in 48 GB or 32 GB, compare the RTX 4090 48G and RTX 5090, or read the RTX 4090 vs RTX 5090 vs A100 comparison.

How it works

  1. Sign up with an email address or a Google account.
  2. Add balance. Visa or Mastercard, Alipay (in most countries) or Google Pay / Apple Pay in one checkout, USDT on TRC-20, or bank transfer for business customers. See payments.
  3. Deploy. On Deploy, choose a template (Ollama, JupyterLab + Ollama, ComfyUI, LLaMA-Factory) or Raw GPU, and the A100 80G.
  4. Work. About 2 minutes later the instance row shows root SSH and JupyterLab; API templates also get a public HTTPS URL.
  5. Destroy. Copy off what you want to keep, then destroy the instance. Billing stops and unused minutes go back to your balance.

Billing in plain words

  • Billed per minute of GPU time, with a 1-minute minimum, so a 20-minute test is billed as 20 minutes.
  • While an instance runs, each started hour is pre-charged from your balance; unused minutes go back to your balance when you destroy the instance.
  • No subscription and no monthly fee; the balance does not expire.
  • Persistent storage across restarts is still on the roadmap, so copy checkpoints and outputs off before you destroy.

FAQ

Is the A100 80G billed by the hour? No. It is billed per minute of GPU time, with a 1-minute minimum. Each started hour is pre-charged from your prepaid balance, and unused minutes go back to your balance when you destroy the instance.

What is the A100 80G good for? Jobs that need more memory on one GPU than consumer cards offer. Which model fits depends on its size, quantization and context length; see the guide on how much VRAM an LLM needs.

Does vLLM support the A100? Yes. vLLM’s documentation lists GPUs with compute capability 7.5 or higher and names the A100 as an example.

How fast does it start? An instance usually takes about 2 minutes from Deploy to a running machine.

Where is the current rate? On the pricing page, which shows the live rate for every card.

Ready to start?

Deploy an A100 80G · See the current rate

Sources

All read on 24 October 2026 (Beijing time, UTC+8).