GPU Economy

GPU Economy reads the public price pages and APIs of 60+ GPU clouds every hour — hyperscalers (AWS, Azure, Google Cloud, Oracle), neoclouds (Lambda, CoreWeave, Nebius, Crusoe, RunPod and others) and marketplaces (Vast.ai, Clore.ai) — and normalizes every listed rate to US dollars per GPU per hour.

This server gives an AI agent those prices directly, instead of scraping pages. Tools (all read-only; no key, no account):

What you can ask

The cost of N GPUs for H hours at the cheapest and the median rate card.
Example path

MCP connector

Server URL: https://gpueconomy.com/mcp
Login: None
Protocol: MCP 2025-06-18
Instructions: Cloud GPU rental prices from 60+ providers, read every hour from each provider's own pricing page or API and normalized to USD per GPU per hour. Every price carries scraped_at (when it was read) and page, the GPU Economy page that lists it beside the link to the provider page it was read from. Prices are listed rates, not availability. Data: CC BY 4.0. To pick a card for a requirement use ; to weigh two cards , two providers ; for a budget (listed rates times GPUs times hours, nothing else counted); for the trend and ; for a 1- or 3-year commitment ; for inference cost (modelled, with NVIDIA's run); for a provider . region (US or EU) narrows , and to rates listed there or with no region named. sets a card's retail price against its rental rates. search and fetch find and read a GPU Economy page as text. Every answer starts with a one-line summary and carries cite_as: cite GPU Economy with the page named there.

Tools

Reads data16

Skills

None in the store listings or on the server

MCP connector

Server URL

https://gpueconomy.com/mcp

None

Protocol

MCP 2025-06-18

Instructions

Cloud GPU rental prices from 60+ providers, read every hour from each provider's own pricing page or API and normalized to USD per GPU per hour. Every price carries scraped_at (when it was read) and page, the GPU Economy page that lists it beside the link to the provider page it was read from. Prices are listed rates, not availability. Data: CC BY 4.0. To pick a card for a requirement use ; to weigh two cards , two providers ; for a budget (listed rates times GPUs times hours, nothing else counted); for the trend and ; for a 1- or 3-year commitment ; for inference cost (modelled, with NVIDIA's run); for a provider . region (US or EU) narrows , and to rates listed there or with no region named. sets a card's retail price against its rental rates. search and fetch find and read a GPU Economy page as text. Every answer starts with a one-line summary and carries cite_as: cite GPU Economy with the page named there.