About Cheap GPUs

Cheap GPUs tracks GPU rental prices, LLM inference costs, and prediction-market forward curves across a growing list of providers, so you can see the cheapest option at a glance instead of checking every provider's pricing page yourself.

What we track

• GPU rentals — current rates for A100/H100/H200/B200 and more, across vast.ai, Akash, RunPod, Lium, Targon, Vultr, and others.
• LLM pricing — per-token cost across major inference providers, with a route finder for a given model and token count.
• Prediction markets — where Kalshi and Polymarket traders expect GPU compute prices to be weeks or months from now.

How it works

Every provider is queried on its own schedule via public APIs, normalized into a common format, and stored so we can chart price history over time. There's no account, login, or paywall — the data is free to browse, and the underlying endpoints are documented in llms.txt.

Contact

Questions or feedback can be sent to hello at cheap-gpus dot com.