What I like most about this is that it completely removes the headache of infrastructure management for my project. All of that is handled by it, so I don’t need to worry about scaling GPUs; it can handle whatever number of GPUs I need. It offers essentially infinite scaling with zero waste, and I don’t have to set up anything GPU-related myself. Another interesting part is that they built their own file system and container runtime, so I don’t need to worry about those either they manage it on their own. Because of that, I can just focus on my code and building, instead of worrying about deployment and infrastructure. Modal works well with the other tools I use as a developer, especially Python and GitHub. I can keep most of my application code in my existing workflow and. The integration is straightforward, so I do not have to completely change my development setup. This makes it convenient for testing AI workloads, running jobs, and scaling them when needed. Review collected by and hosted on G2.com.

What I dislike is that if I want to change the code, there are some commands I should use, which is great. However, later on, if I want to move the deployment to another server, it requires a lot of changes in my project, and that’s a problem for me. Basically, you become dependent on this model, so if they raise the price later, it becomes difficult to change the deployment server. Also, since it doesn’t always run when no one is using it and it goes into sleep mode, when a user suddenly tries to access it, it takes some time to load, which can make customers unhappy. Review collected by and hosted on G2.com.