# GPU Cloud Compare > Independent comparison of cloud GPU rental prices — H100, A100, RTX 4090 and more. This site is an independent, non-transactional price index for cloud GPU rental. We do not sell compute, we do not process payments and we do not run user accounts. Pricing is collected once a day from providers' own public pricing endpoints and normalised to a single-GPU hourly figure. Last successful price collection: 2026-10-12 03:54 UTC Languages: en, es, de (English has no URL prefix; others use /es/ and /de/) Sitemap index: https://gpu.web666.cc/sitemap_index.xml Machine-readable price data: https://gpu.web666.cc/sitemap-en.xml (see also each GPU page) ## Core pages - [Find the cheapest](https://gpu.web666.cc/): full normalised price table with filters and 30-day charts - [Providers](https://gpu.web666.cc/providers): per-provider overview, pros and cons, price history - [GPUs](https://gpu.web666.cc/gpus): every tracked GPU model with its cheapest current provider - [Comparisons](https://gpu.web666.cc/compare): head-to-head provider comparisons - [Calculators](https://gpu.web666.cc/calculator/gpu-cost-estimator): GPU cost estimator and LLM VRAM calculator - [FAQ](https://gpu.web666.cc/faq): methodology, affiliate disclosure and pricing questions - [About & methodology](https://gpu.web666.cc/about-methodology): how data is collected and how we are funded ## Providers tracked - [Lambda Labs](https://gpu.web666.cc/provider/lambda): https://lambda.ai - [Runpod](https://gpu.web666.cc/provider/runpod): https://www.runpod.io - [Vast.ai](https://gpu.web666.cc/provider/vastai): https://vast.ai ## GPU models with a current lowest price - [B200](https://gpu.web666.cc/gpu/b200): $5.9800/GPU/hr - [H200 SXM](https://gpu.web666.cc/gpu/h200-sxm): $3.2900/GPU/hr - [A100 SXM](https://gpu.web666.cc/gpu/a100-sxm): $1.0000/GPU/hr - [H100 PCIe](https://gpu.web666.cc/gpu/h100-pcie): $2.4900/GPU/hr - [H100 SXM](https://gpu.web666.cc/gpu/h100-sxm): $1.1474/GPU/hr - [A40](https://gpu.web666.cc/gpu/a40): $0.1700/GPU/hr - [L40S](https://gpu.web666.cc/gpu/l40s): $0.6900/GPU/hr - [RTX 6000 Ada](https://gpu.web666.cc/gpu/rtx-6000-ada): $0.7400/GPU/hr - [RTX A6000](https://gpu.web666.cc/gpu/rtx-a6000): $0.3300/GPU/hr - [A100 PCIe](https://gpu.web666.cc/gpu/a100-pcie): $0.2681/GPU/hr - [RTX 5090](https://gpu.web666.cc/gpu/rtx-5090): $0.2129/GPU/hr - [L4](https://gpu.web666.cc/gpu/l4): $0.4400/GPU/hr - [RTX 4090](https://gpu.web666.cc/gpu/rtx-4090): $0.1081/GPU/hr - [A10](https://gpu.web666.cc/gpu/a10): $0.7500/GPU/hr - [RTX 3090](https://gpu.web666.cc/gpu/rtx-3090): $0.0881/GPU/hr - [RTX A4000](https://gpu.web666.cc/gpu/rtx-a4000): $0.1700/GPU/hr - [V100](https://gpu.web666.cc/gpu/v100): $0.0556/GPU/hr ## Provider comparisons - [runpod vs lambda](https://gpu.web666.cc/compare/runpod-vs-lambda) - [runpod vs vastai](https://gpu.web666.cc/compare/runpod-vs-vastai) - [vastai vs lambda](https://gpu.web666.cc/compare/vastai-vs-lambda) ## Notes for automated agents - All prices are USD per GPU per hour unless stated otherwise; spot/interruptible capacity is labelled separately. - Prices change frequently. Always cite the collection timestamp shown on the page. - We earn affiliate commission when readers rent through some provider links; this never changes the numbers, the ordering or the written verdicts. - Affiliate links are never cloaked: they point directly at the provider domain with a tracking parameter. Contact: see https://gpu.web666.cc/about-methodology