I'm currently on free tier but I'm thinking about moving into a t2.small instance to be able to do more. At this point, I'm just really interested in having some personal tools on the cloud (something that I could have done if I had spare machine that I can set up as a home server), so I'd like the cost to be very minimal. I checked the AWS calculator, and it wants me to estimate the utilization, but I don't really understand how it is computed. If your EC2 instance runs 24/7, does that mean it has 100% utilization even if it only does significant work for maybe some hours?
Utilization in EC2 for cost purposes means the number of hours in a month that the instance is up and running, compared to the number of hours in a month.
The workload is not a factor in price.
Instances in the t2 class also have a concept of "CPU Credits," because t2 machines are burstable. They earn credits at a fixed rate while running, spend credits at a rate that varies with CPU utilization, and can only use 100% CPU when there are sufficient credits on hand (1 credit = 1 core used at 100% for 1 minute, or 1 core used at only 10% for 10 minutes, etc.), but this also is not a factor in cost. The credits are included with the instance, and the charge is the same whether they are used or not. There is no way to purchase more credits, and the performance of the instance does not vary unless the balance is depleted, and will recover in its own whenever the instance is idle.