Paper for MC 1.14

Minecraft 1.14.1

Minecraft 1.14.1 was pushed out relatively quickly to fix some of the most glaring issues with Mojang’s 1.14 release. One of these changes resulted in a memory leak in early versions of Paper 1.14.1 but that has since been fixed.

1.14.1 is a step in the right direction but clearly hasn’t resolved all the issues in the game. There are still gameplay related bugs and some concerns about chunk I/O performance. Mojang has already announced a 1.14.2 update so we’re now recommending those upgrading from older versions wait.

Some users have already chosen to update to 1.14 and 1.14.1 despite warnings. That’s fine, but it’s clearly not ready for those of you looking for a bug-free experience. It is not behaving as nicely as 1.13.2 at the moment. Those of you who updated need to be staying up-to-date. Bugs get fixed, if you aren’t on the latest build expect your reports to be closed.

On Paper’s side there are still patches that need to be re-implemented and tested. We’re also constantly finding new issues from Vanilla that people request fixes for and we will probably start looking at doing some of the most critical ones soon.

For those who haven’t read the original post yet it is still below and I highly recommend reading through it.


Minecraft 1.14

Minecraft 1.14 released with new content and many system changes under the hood. Internally, this was actually a very large update. Many systems have changed. Chunk management, lighting, entity tracking, and a whole bunch more have been changed. Many of these changes seem to be technical improvements over the existing systems. Unfortunately, it appears that many of these systems were not thoroughly tested before release. This is not even factoring in all the typical “new release” teething problems that occur as features and APIs are updated. As a result, 1.14 is not a recommended update at this time.

We recommend all servers wait until Minecraft 1.14.1 is available before even considering updating or deploying to production.

We are largely focusing on getting all our existing patches tested and stable during these first few weeks and while we may fix the occasional vanilla issue, it is unlikely we will fix all the issues plaguing vanilla 1.14. Which is why we recommend skipping MC 1.14 and waiting for 1.14.1.

@electronicboy took the lead on this update and received significant help from @Spottedleaf and other contributors. Feel free to thank and throw tacos at them.

We will provide updates and recommendations on this post and on Twitter as we get to the MC 1.14.1 release which is expected to be soon.

The downloads page on the main site will not be updated until we feel 1.14 is in a better place. Downloads for these new, unstable, development releases of 1.14 will be available on the downloads API mentioned in this post.

If you choose to run 1.14 you need to know you will experience problems. If you report those you need to do your due diligence. Does it occur on Vanilla? Does it occur on CraftBukkit? Does it occur on Spigot? The team is already going to be very busy trying to fix and improve the software. Bad bug reports just slow everyone down. This release is being treated as a testing/development release. We expect you to actually try with your reports. Failure to do so may result in your report being closed.

Paper 1.12.2 Support

As many of you know, we have been supporting the Paper 1.12.2 release with applicable bug fixes as we find them. Paper 1.12.2 has been feature-frozen for some time now. The decision was made to continue supporting 1.12.2 because we were not happy with the overall state of 1.13 at release, and while it improved it never got where we thought it should be.

We had hoped for a smoother 1.14 release. For a while there it looked promising. Obviously, that is not what we ended up with.

Luckily, 1.12.2 has proven to be a very solid release and we do not get many tickets from it anymore. As this release cycle goes forward and 1.14.1 releases and is polished you can probably safely expect 1.12.2 to slowly lose support from us. That being said, it is a stable release and you should not rush to replace it with anything you are not comfortable with. We hope that over time 1.14(.1?) will be a proper successor.

Paper 1.13.2 Support

The 1.14 update and the above 1.12.2 support puts 1.13.2 in a weird place. Many of you are running 1.13.2 right now. You should not update yet. However, it is unlikely that we have the means or the time to continue providing it with feature updates. We may work on significant bug fixes for 1.13.2 but you should expect our focus to move to 1.14.

Our goal is to get 1.14.1 (or later just depending on how long this all takes to sort out) polished and at least at 1.13.2’s level. Once that occurs, we will likely drop 1.13.2 support entirely but not before 1.14 is stable, ready, and recommended. You have time.

The long-term goals right now are to polish the 1.14 series of releases enough for users currently on 1.13.2 to make a quick and easy upgrade. Once that’s done, we’d like to continue polishing such that 1.12.2 users can make a slower and more long term upgrade. All of this is subject to change however based on how long this all takes and what happens with vanilla and upstream releases.

1.14 is not ready and we do not recommend you update yet. We will update you as the situation changes.


There are API breakages from upstream in this update. It is not as bad as 1.13 was but there are notable breakages that you should begin looking at and figuring out what you want to start supporting and when.

This update does not require updating api-version in your plugin.yml files. api-version: 1.13 will work just fine. You can optionally bump it to api-version: 1.14 if you wish, however it will make your plugin no longer function on 1.13 servers.

This section will be updated as we get questions and if we notice anything else not being covered elsewhere.

You should also check upstream’s release announcement for other developer-relevant information.


1.14 is now available. 1.14 is not ready. We do not recommend updating your servers yet. We will alert you when this changes.

Thank you for all the support over the years. A special thank you to our contributors as well.

We are a small team of volunteers doing our best and we hope that we are living up to all the support we get from all of you.



You can join our Discord ( and/or follow us on Twitter ( to receive alerts as we go forward with 1.14 and the project in general.



You guys are awesome. Thanks for the update on the flow of things.

I just want to say thanks to everyone on the PaperMC project - your work makes life on earth better for all of us.

Keep it up, and a big, big thank you for everything you have done and continue to do.


I read a lot of positive options for Paper, that it is 100% compatible with spigot plugins, so it seems, and that I have nothing to do to upgrade from Spigot to Paper… NOTHING!!
It did not say that the server does nothing afterwards??? So, only changed spigot.jar with paper.jar and the server eats up all processor utilization and does not progress… Changing it back to Spigot crashes the server even more… Taking a backup from last month, sorry to lose a bit, but this hub im trying to upgrade is rather small and very simple. That restored version works with Spigot but not with Paper…
How is this possible?
If I look at my server now, I only see a white-hot server and no progress…
I have no idea what to do except trust Spigot only… I’m hoping you guys can make me see another way…

Thanks a lot for replying to me on this matter…

I did a pre-forceupgrade on this world, so Spigot ordered it in case to go to 1.14…

Hi. We are trying to run a 1.14.2 server for two local people on our lan. I’m running it on our surface 4 pro with these specs:

Intel ® Core™ i7-6650u cpu @2.20Ghz
16 GB ram

We are seeing can’t keep up messages from time to time, and hitching in the game.

My question is if the surface is under powered for being a server for us, or is it just the current state of 1.14 development of paper causing issues?

Is spiggot a better 1.14.2 choice for now?

That’s 1.14 as a whole on an underpowered mobile CPU, ditching paper would mean that you’d lose a few pretty critical world-gen crash fixes as well as boosts such as incremental chunk saving (vs a large tps loss spike every autosave interval)

The I7 surface was not the slow bottom line surface… it was the top of the line fastest one available. I can even run zbrush and do digital sculpting on it using millions of polys.

I did have distance turned up to 18, and turning it down to 15 seems to have corrected the problem. (Distance of 12 is recommended for 1.14, but that seems like a really low distance, since bedrock does 30.)

We are also running the latest 1.14 test release of optifine that increased chunk loading speed 2x on the latest r13 build, so that might be helping the clients.

Bedrock edition can do up to 30 chunks of “render distance”, but what you’re not seeing is that very few of those chunks are actually ticking at all whatsoever. They’re basically just there for players to have a pretty view. You should set your render distance to 12 at maximum, there should be no need to set it higher for basically any reason. If players want to have the same effect of bedrock edition’s 30 chunks of ““render distance”” then they can increase their own render distance, as the client will keep a cache of chunks they’ve visited as they explore, even if those chunks aren’t loaded by the server still.

That good to know! BTW, something we noticed yesterday… my framerates with optifine and sildur’s re usuall 50-60… after a few hours that had dropped to 20-30 for some reason.

I was killed probably 2 chunks from my bed, and when I revived, I was back up to 60 fps!

Any idea why that would be happening?

(I’ve run servers for us on other games, but have only been running a minecraft server for only a few months.)

Thanks for that tip!

Any ETA when PaperMC for 1.14.x can be called „stable“?

Depends on your notation of stable, 1.14.x is stable in that it’s not going to crash on you (We’re somewhat reasonably confident of that), that’s not to say that there isn’t large scale performance issues all the way down from vanilla which are hurting servers harshly; Shamefully, the scope of some of the issues involved is large making patching them before mojang super costly in terms of time

1 Like

Thanks for the answer. Since the players on my server are bugging me: Do you think it is save to update without plugins with no more than 6 players online at the same time? I read some threads about ram problems with 1.14.3 here. Or is this only related to plugins? I don’t want to come back here after the update from 1.13.2 to 1.14.3 and cry my heart out and you say “well, we did not say it is stable. so…” :wink:

It’s stable, just literally nobody is going to recommend that you update it or performance issues are pretty much guaranteed given the state of vanilla, most memory issues are really just people not understanding how javas memory work, plugins are also a common major contributor to leaks, you can’t really say what’s going on without taking and looking into the proper reports, e.g. heap dumps, will show you exactly what is taking up space in the heap