Running in ~30 s
Pre-cached image, ports mapped, TLS terminated — the stack is working before you finish reading this.
Template · LLM serving & chat · CUDA 13
Run vLLM on a cloud GPU, in 30 seconds
Production LLM serving with PagedAttention — an OpenAI-compatible endpoint from any HF model. vLLM is the production standard for serving open LLMs: PagedAttention, continuous batching, tensor parallelism and an OpenAI-compatible API. The template exposes port 8000 with TLS, reads MODEL from an environment variable and caches weights on the instance disk or a mounted volume. From $0.159/hr on an interruptible RTX 5090.
Template is free
You pay the GPU price only: vLLM on a RTX 5090 is $0.318/hr on-demand, $0.159/hr interruptible, billed per second.
Volumes for state
Models, datasets and outputs live on a $0.08/GB/mo volume; the instance stays disposable.
Private by default
Dedicated GPU, encrypted disk, crypto payments, no KYC, no stored IPs — Jupyter and SSH behind your own credentials.
Best GPUs for vLLM
Three price points that run this template well — Good, Better, Best. Every model in the catalogue works; these are the value picks.
| Tier | GPU | VRAM | On-demand | Interruptible | Why this card | Action |
|---|---|---|---|---|---|---|
| Good | RTX 5090 | 32 GB | $0.318 | $0.159 | 32 GB GDDR7: the lowest $/token for 7B–14B FP16 and 32B 4-bit models. | Deploy |
| Better | L40S | 48 GB | $0.466 | $0.233 | 48 GB and datacenter cooling for 24/7 endpoints up to 32B. | Deploy |
| Best | H100 PCIE | 80 GB | $2.147 | $1.073 | 80 GB for quantized 70B on one card, FP8 for maximum throughput. | Deploy |
Need more VRAM? The full catalogue lists all 76 models with live availability; the VRAM guide sizes models to cards.
Deploy vLLM from the console, CLI or API
Pick the template in the console deploy bar, or script it:
- Console — filter by GPU, choose vLLM in the template picker, set disk and env, deploy.
- CLI — pip install powergpu, then the command on the right. CLI reference.
- API — POST /v1/instances with "template": "vllm". REST reference.
- Own image — any OCI reference works too; we inject the NVIDIA runtime. Template docs.
$ powergpu launch --gpu rtx-5090 --template vllm \
--disk 100 --volume models:/workspace/models
✓ instance i-7a41c0e2 running (27.9s)
# vLLM · RTX 5090 · $0.318/hr · per second
# https://i-7a41c0e2.powergpu.io:8888 (TLS)
$ powergpu stop i-7a41c0e2 # billing ends this secondWhat is inside the vLLM template
| Image | powergpu/vllm |
|---|---|
| CUDA | CUDA 13 |
| Access | also builds for ARM hosts · SSH shell · JupyterLab on a mapped port |
| Category | LLM serving & chat |
| Storage | Instance NVMe disk (sized at deploy) + optional network volumes |
| Billing | GPU price only, per second — no template fee, no setup fee |
Other llm serving & chat templates
vLLM Omni on a cloud GPU — vLLM extended for multimodal models — vision and audio inputs on the same fast server.
SGLang on a cloud GPU — High-throughput serving with RadixAttention — excels at structured and agentic workloads.
Llama.cpp on a cloud GPU — GGUF inference with a built-in server — the lightest way to run quantized models.
Ollama on a cloud GPU — Pull and run quantized models in one command, with a clean REST API.
Open WebUI (Ollama) on a cloud GPU — A full chat UI over Ollama — conversations, RAG and model management in the browser.
Oobabooga Text Gen on a cloud GPU — The text-generation WebUI & API — load GPTQ, EXL2 or GGUF with extensions.- All 37 templates
vLLM on a cloud GPU: FAQ
Environment variables, ports and custom images are covered in the template docs.
How much does it cost to run vLLM on a cloud GPU?
Only the GPU price — the template is free. From $0.159/hr on an interruptible RTX 5090, $0.318/hr on-demand on a RTX 5090. Billing is per second, so an hour of tinkering costs an hour, not a day. Storage is $0.08/GB/month.
How long does vLLM take to start?
About 30 seconds from the deploy click: the image is pre-cached on hosts, ports are mapped and TLS is terminated for you. Restarting a stopped instance is faster, and your disk is exactly as you left it.
Can I keep my models and outputs between sessions?
Yes — attach a volume at deploy. Everything on the volume survives instance destruction and mounts on the next instance in the region in seconds, at $0.08/GB/month. The instance disk itself survives stop/start but not destroy.
Is the endpoint really OpenAI-compatible?
Yes — point any OpenAI SDK at https://<instance>.powergpu.io:8000/v1 with the API key you set in VLLM_API_KEY. Chat, completions, embeddings and streaming all work unchanged.