A Node.js API is one of the cheapest workloads to host \u2014 a typical Express or Fastify service idles at 80\u2013150 MB of RAM and handles a few hundred requests per second on a single core. That means a $6\/month VPS with 1 GB RAM runs a real production API with room to spare. The catch is the build step: Next.js, Nuxt, and NestJS builds routinely spike to 1.5\u20132.5 GB of memory for a minute or two, which is why so many “Node on a $5 VPS” tutorials end with an OOM-killed build process. Knowing which numbers matter \u2014 runtime memory, build memory, and concurrent connections \u2014 turns a $6 box from a toy into a legitimate deployment target.<\/p>\n\n\n\n
The decision starts with your stack. A plain Express or Fastify API fits on 512 MB\u20131 GB. A server-rendered Next.js app with a small database fits on 1\u20132 GB as long as you build locally or add swap. A full monorepo with multiple services, a queue worker, and Redis needs 2 GB+. When you compare budget VPS plans side by side<\/a>, the useful question is not “how many cores” but “does the plan have enough RAM for my build spike and enough disk for my node_modules.”<\/p>\n\n\n\nReal Memory Profiles for Common Node Workloads<\/h2>\n\n\n\n