What is the difference between serverless GPU and instance pricing?

With a rented instance you control the machine and pay a flat hourly rate whether it is busy or not. That is efficient when the GPU is loaded most of the time and predictable when you need a persistent environment, but you eat the cost of every idle minute.

Serverless platforms wrap the GPU behind a function or endpoint and charge per second of execution, often with a per-request component. You pay nothing when no requests arrive, which is ideal for bursty traffic, occasional batch jobs, or a demo that sits quiet most of the day. The tradeoffs are cold-start latency when a worker spins up and a higher effective per-second rate.

The rule of thumb: if your duty cycle is low, serverless usually costs less overall; if you would keep an instance busy for hours at a stretch, the dedicated instance is cheaper. Estimate your requests per day and their duration to see which side you fall on.