It varies wildly, even given the runtime (5 minutes) I could easily see a couple of orders magnitude change depending. A lot based on the amount of work, but also quality, seniority of the artists, etc. One way to intuit this is to look at movies: Compare e.g. Avatar to Blair Witch Project. The difference in running time doesn’t reflect the difference in budget. Even in animation, there are big differences in budget.
Day rates for artists vary a lot, I’ve seen anything from 200 to 700, but you might go lower (or higher) based on your circumstances.
Finally as a working artist, I have to say if you go below a certain minimum pay, you are basically getting your artists to subsidize your production. I think this is fair for a non commercial ‘labor of love’ or better, open movie (elephants dream, BBB, etc.) but not for a commercial project (sausage party) In the former case, it’s artists with a like-minded desire collaborating together, in the later, it’s wage theft.
I’m also willing to bet that most of the working artists in this thread (hi folks) won’t be too forthcoming even though they know full well how to budget stuff. It’s no fun being held to a low rate you did when you were green, or conversely, not being considered for a job you wanted because you were deemed ‘too expensive’ based on a high budget thing you did.
So that being said, here’s how you could do it:
If you have a script, boards, etc., break down the short into tasks. Really work through everything to get every single shot, the edit, and sound done.
Use a ‘standard’ theoretical person to estimate the time on these tasks - I often do ‘reasonably competent professional’ which gives me a day rate of around 300-500 here. If you can’t judge that yet, you can estimate how long it would take you to do everything, and how much you’d pay yourself to do it (your day rate)
Then you just do math. Typical people use spreadsheets, Some use dedicated planning software (like planner or microsoft project) which I find to be a bit cumbersome. Personally I use blender’s node editor in combination with spreadsheets (I have an addon called orgnode that represents tasks as individual nodes in the editor. This also allows me to plan ‘what gets done first’).
If the project is not your own, a certain amount of guessing comes into play, and even if it is yours, I also recommend multiplying everything by 2 or 3 at the end (seriously) as even with experience, things ‘blow up’. This last advice was from working at an engineering firm who used to do that - we’d end up making 10-15% profit on our jobs in the end. (would have lost our shirts otherwise)