I made a clock program for the ti 84 plus ce where it uses the calculators clock but I noticed the clock always is way off even when I set it multiple times.

The time is in ram, the clock resets every time your ram resets.

What kind of drift are you seeing? Is it going all the way back to 2015, or drifting by a few seconds each day, or something more dramatic than that?

OK so I did know that about the ram but I'll have to check about how far it drifts it is around a couple hours off or so after one night the day had been fine though. I'll do a test and set it then see what it is in the morning. Does turning the calculator off affect it? It never ran dead.

No, turning the calculator off shouldn't matter. Does your clock program rely on startTmr and checkTmr or some other clock commands?
If it does, LogicalJoe found a bug in the OS last year that was added around version 5.4.0 which causes the clock to be off by a round number of days depending on the month (because the number of days in each month got shifted by 1 month)
The only fix for now is to correct the time by accounting for this bug (removing or adding days depending on which month it is)
That being said, if you're really talking about "a couple hours", then its not that. I wouldn't put it past TI to have introduced even more timer bugs though. Its not like the gregorian calendar has changed since 2003 when that code was initially written

It could be the big your talking about because sometimes it does it and sometimes not I looked at it this morning and it was still accurate. I do use the get time command and output the answer.

Also I don't think it's the program. Also why are consumers better at programming and using they're products better then themselves. Ti is idiots.

making the TI-84 Plus clock the most useless and most unreliable clock in existence. What the hell were they thinking?

Ok so I did a test and it was 35 hours ahead in a span of 30 hours.