Pycharm 2025.3 - badly broken

After uninstalling Pycharm, removing manually all Pycharm specific directories I found on my drive, like:

- `~/var/app/com.jetbrains.PyCharm-Professional`

- `~/.cache/JetBrains`

- `~/.config/JetBrains/`

- `~/.local/share/JetBrains/`

I've installed Pycharm again it and it seems to work correctly. Seems, because I observe it freezing sometimes for no obvious reason. Freezing means it's unresponsive and there's a huge CPU usage. This happened to me for example when I tried to add an external tool config, or when setting up remote (docker compose) python interpreter.

Also it has issues with resolving imports. Remote interpreter is setup but imports are marked as invalid “Cannot resolve imported module `structlog`”. However ctrl+click properly takes me to the target file.

Note: I use PopOs.

Hello, Azgertis ,

Please check if the issue with freezes is still present if disable all non-bundled plugins (Settings | Plugins | click the cog icon | Disable All Downloaded Plugins) and File | Invalidate Caches with all options toggled.

If it does not help, please create YouTrack ticket and attach a log archive, which can be found under Help | Collect Logs and Diagnostic Data. Also, please create and share a CPU snapshot by following this article (Section: The CPU usage is high or IDE is slow).

I have the same issue and I'm on PopOs too. I just did the update to 2025.3 and got that issue with the java process consuming around 65% CPU and 24 Go of RAM.

At the moment:

1. Pycharm 2025.3 works mostly well. CPU usage is not bad - seems to be normal for Pycharm - when idle it jumps in the range of 5-40%.

2. Turning off RyeCharm solved issues with imports and usage informations

3. I can't make Ruff to work on save and do code formatting for example. RyeCharm did that well but caused other issues.

4. There are some strange inspection errors that come from IDK where (shown even when all inspections in ruff etc are off), like “`MyClass` object has no attribute `data`”, while it has it.

But again - I had to uninstall Pycharm and manuall remove all related folders to make it work.

I had my PyCharm limited to 2 GB of RAM and it worked perfectly.
But two days ago I updated PyCharm to version 2025.3, and it suddenly became extremely slow and freezes my entire Mac.

I increased the RAM limit to 8 GB, but it still keeps freezing.
It’s impossible to work with this version of the IDE — everything is completely stuck for hours.
Right now, I can’t even open or change the settings because it’s so slow.

I'm on Ubuntu 24.04, and 2025.3 can't even startup, it freezes when loading the project. It always had 8gb…all this for rounded corners? (sorry ;)

I'm back on 2025.2. I can understand that a product needs to evolve. But this looks like an very shaky release!

I've found that on Fedora 41the process actually goes into the OS Stopped state and needs a `kill -SIGCONT` (several times) for the process to complete startup. Once you've done that and it's completed the startup it appears to function properly.

Ubuntu 24.04 - I also get it going into stopped state (htop state is `T` - recoverable by kill -SIGCONT) every time I create or open an python venv. 

The logs show `pycharm/plugins/python-ce/helpers/generator3/__main__.py` trying to import my projects code and getting many ModuleNotFoundErrors (presumably since it's a new venv) - finally “Fatal Python error: Segmentation fault”.

Unfortunately I can't share the full logs, but it seems this generator3 is causing the crash

edit - the freezes additionally happen on opening the project and don't produce the above segfault,

Here are some extracted logs, annotated with where the freeze occurred and I send SIGCONT

```
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.d.i.f.FrontendXBreakpointManager - Breakpoint created for XBreakpointId(uid=1)
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.d.i.f.FrontendXBreakpointManager - Breakpoint created for XBreakpointId(uid=2)
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.d.i.f.FrontendXBreakpointManager - Breakpoint created for XBreakpointId(uid=3)
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.d.i.f.FrontendXBreakpointManager - Breakpoint created for XBreakpointId(uid=4)
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.d.i.f.FrontendXBreakpointManager - Breakpoint created for XBreakpointId(uid=5)
2025-12-12 13:25:09,496 [   4147]   INFO - #c.i.p.m.f.i.ImportsRankingModelService - Successfully loaded imports ranking model
2025-12-12 13:25:09,502 [   4153]   INFO - #git4idea.commands.GitHandler - git version 2.43.0
2025-12-12 13:25:09,507 [   4158]   INFO - #c.i.v.l.d.VcsLogData - Vcs log index is turned off in the registry
2025-12-12 13:25:09,507 [   4158]   INFO - #git4idea.config.GitExecutableFileTester - Git version for /usr/bin/git: 2.43.0.0 (UNIX)
 

 FREEZES HERE

2025-12-12 13:25:51,679 [  46330]   WARN - #c.i.u.ShellEnvironmentReader - shell env loader is timed out
2025-12-12 13:25:51,684 [  46335]   INFO - #c.i.m.l.c.q.QuotaManager2Impl - Updating quota refill state
2025-12-12 13:25:51,685 [  46336]   INFO - #c.i.m.l.c.q.QuotaManager2Impl - New quota refill state is: Error(exception=com.intellij.ml.llm.core.quota.ResultDoesNotMatchConditionException: Returned result does not match the condition, previous=null)
2025-12-12 13:25:51,691 [  46342]   INFO - #c.i.w.i.i.j.s.JpsGlobalModelSynchronizerImpl - Loading global entities com.intellij.platform.workspace.jps.entities.SdkEntity from files
2025-12-12 13:25:51,762 [  46413]   WARN - #c.i.m.l.c.m.LlmProfileRulesLoader - Failed to load cloud rules: Remote host terminated the handshake
javax.net.ssl.SSLHandshakeException: Remote host terminated the handshake
    at java.base/sun.security.ssl.SSLSocketImpl.handleEOF(SSLSocketImpl.java:1714)
    at java.base/sun.security.ssl.SSLSocketImpl.decode(SSLSocketImpl.java:1514)
    at java.base/sun.security.ssl.SSLSocketImpl.readHandshakeRecord(SSLSocketImpl.java:1421)
    at java.base/sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:455)
    at java.base/sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:426)
    at java.base/sun.net.www.protocol.https.HttpsClient.afterConnect(HttpsClient.java:586)
    at java.base/sun.net.www.protocol.https.AbstractDelegateHttpsURLConnection.connect(AbstractDelegateHttpsURLConnection.java:187)
    at java.base/sun.net.www.protocol.http.HttpURLConnection.getInputStream0(HttpURLConnection.java:1691)
    at java.base/sun.net.www.protocol.http.HttpURLConnection.getInputStream(HttpURLConnection.java:1615)
    at java.base/java.net.HttpURLConnection.getResponseCode(HttpURLConnection.java:531)
    at java.base/sun.net.www.protocol.https.HttpsURLConnectionImpl.getResponseCode(HttpsURLConnectionImpl.java:307)
    at com.intellij.util.io.HttpRequests.openConnection(HttpRequests.java:641)
    at com.intellij.util.io.HttpRequests$RequestImpl.getConnection(HttpRequests.java:381)
    at com.intellij.util.io.HttpRequests$RequestImpl.getInputStream(HttpRequests.java:389)
    at com.intellij.ml.llm.core.models.LlmProfileRulesLoader.loadFromUrl$lambda$0(LlmProfileRulesLoader.kt:198)
    at com.intellij.util.io.HttpRequests.doProcess(HttpRequests.java:544)
    at com.intellij.util.io.HttpRequests.process(HttpRequests.java:526)
    at com.intellij.util.io.HttpRequests$RequestBuilderImpl.connect(HttpRequests.java:355)
    at com.intellij.ml.llm.core.models.LlmProfileRulesLoader.loadFromUrl(LlmProfileRulesLoader.kt:197)
    at com.intellij.ml.llm.core.models.LlmProfileRulesLoader.loadCloudRules(LlmProfileRulesLoader.kt:177)
    at com.intellij.ml.llm.core.models.LlmProfileRulesLoader.access$loadCloudRules(LlmProfileRulesLoader.kt:42)
    at com.intellij.ml.llm.core.models.LlmProfileRulesLoader$initiateLoading$2.invokeSuspend(LlmProfileRulesLoader.kt:97)
    at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:34)
    at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:100)
    at kotlinx.coroutines.internal.SoftLimitedDispatcher$Worker.run(SoftLimitedDispatcher.kt:130)
    at kotlinx.coroutines.scheduling.TaskImpl.run(Tasks.kt:89)
    at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:610)
    at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.executeTask(CoroutineScheduler.kt:1188)
    at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.runWorker(CoroutineScheduler.kt:775)
    at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:762)
    Suppressed: java.net.SocketException: Broken pipe
        at java.base/sun.nio.ch.SocketDispatcher.write0(Native Method)
        at java.base/sun.nio.ch.SocketDispatcher.write(SocketDispatcher.java:62)
        at java.base/sun.nio.ch.NioSocketImpl.tryWrite(NioSocketImpl.java:394)
        at java.base/sun.nio.ch.NioSocketImpl.implWrite(NioSocketImpl.java:410)
        at java.base/sun.nio.ch.NioSocketImpl.write(NioSocketImpl.java:440)
        at java.base/sun.nio.ch.NioSocketImpl$2.write(NioSocketImpl.java:819)
        at java.base/java.net.Socket$SocketOutputStream.write(Socket.java:1195)
        at java.base/sun.security.ssl.SSLSocketOutputRecord.encodeAlert(SSLSocketOutputRecord.java:81)
        at java.base/sun.security.ssl.TransportContext.fatal(TransportContext.java:419)
        at java.base/sun.security.ssl.TransportContext.fatal(TransportContext.java:326)
        at java.base/sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:467)
        ... 26 more
Caused by: java.io.EOFException: SSL peer shut down incorrectly
    at java.base/sun.security.ssl.SSLSocketInputRecord.read(SSLSocketInputRecord.java:494)
    at java.base/sun.security.ssl.SSLSocketInputRecord.readHeader(SSLSocketInputRecord.java:483)
    at java.base/sun.security.ssl.SSLSocketInputRecord.decode(SSLSocketInputRecord.java:160)
    at java.base/sun.security.ssl.SSLTransport.decode(SSLTransport.java:111)
    at java.base/sun.security.ssl.SSLSocketImpl.decode(SSLSocketImpl.java:1506)
    ... 28 more
2025-12-12 13:25:51,766 [  46417]   INFO - #c.i.w.i.i.j.s.JpsGlobalModelSynchronizerImpl - Loading global entities com.intellij.platform.workspace.jps.entities.LibraryEntity from files
2025-12-12 13:25:51,799 [  46450]   WARN - #c.i.d.PerformanceWatcherImpl - UI was frozen for 42283ms, details saved to /home/jonathan/.cache/JetBrains/PyCharm2025.3/log/threadDumps-freeze-20251212-132551-PY-253.28294.336-IterationState.getHighlighterEnd-42sec

FREEZES HERE


025-12-12 13:25:51,806 [  46457]   INFO - #c.i.w.i.i.EntitiesOrphanageImpl - Update orphanage. 0 modules added
2025-12-12 13:25:51,808 [  46459]   INFO - #c.i.w.i.i.WorkspaceModelImpl - Project model updated to version 3 in 4 ms: Sync global entities with project: [redacted]
2025-12-12 13:25:51,808 [  46459]   INFO - #c.i.w.i.i.GlobalWorkspaceModel - Global model updated to version 2 in 8 ms: Sync global entities with state
2025-12-12 13:25:51,814 [  46465]   INFO - #c.i.j.d.b.r.HostToClientMessagingService - New subscription for ClientId(value=Host)
2025-12-12 13:26:56,713 [ 111364]   WARN - #c.i.u.ShellEnvironmentReader - shell env loader is timed out
2025-12-12 13:26:56,713 [ 111364]   WARN - #c.i.u.ShellEnvironmentReader - failed to terminate shell env loader process gracefully, terminating forcibly
2025-12-12 13:26:56,715 [ 111366]   WARN - #c.j.p.r.PyVirtualEnvReader - Couldn't read shell environment: [/bin/bash, -l, -i, -c, . '/home/jonathan/code/[redacted]/.venv/bin/activate' && /usr/bin/env -0 > '/tmp/ij-shell-env-data.4223221459284995187.tmp', /home/jonathan/code/[redacted]/.venv/bin/activate], ec=137
2025-12-12 13:26:56,715 [ 111366]  FINER - #c.i.m.l.a.a.c.AcpAgentsConfigService - ACP configuration file watcher enabled. Watching for '/home/jonathan/.jetbrains/acp.json'
2025-12-12 13:26:56,716 [ 111367]   INFO - #c.i.m.l.a.a.r.AcpAgentRegistry - Reloading dynamic ACP agents
2025-12-12 13:26:56,716 [ 111367]   INFO - #c.i.u.g.s.GistStorageImpl - Cleaning old huge-gists dirs from [/home/jonathan/.cache/JetBrains/PyCharm2025.3/caches/huge-gists] ...
2025-12-12 13:26:56,716 [ 111367]   INFO - #c.i.m.l.a.a.r.AcpAgentRegistry - Dynamic ACP agents reloaded: 0 active
2025-12-12 13:26:56,720 [ 111371]   INFO - #c.i.m.l.c.q.QuotaManager2Impl - Updating quota refill state
2025-12-12 13:26:56,720 [ 111371]   INFO - #c.i.m.l.c.q.QuotaManager2Impl - New quota refill state is: Error(exception=com.intellij.ml.llm.core.quota.ResultDoesNotMatchConditionException: Returned result does not match the condition, previous=null)
2025-12-12 13:26:56,721 [ 111372]   INFO - #c.i.j.d.f.r.HostRequestListenerActivity - Connected to host
2025-12-12 13:26:56,722 [ 111373]  FINER - #c.i.m.l.a.a.c.AcpAgentsConfigService - Awaiting for document changes flow...
2025-12-12 13:26:56,723 [ 111374]   INFO - #c.i.u.g.s.GistStorageImpl - Cleaning old huge-gists dirs finished
2025-12-12 13:26:56,726 [ 111377]   INFO - #git4idea.commands.GitHandler - git config took 107218 ms. Command parameters: 
/usr/bin/git -c core.quotepath=false -c log.showSignature=false config --null --get user.name
2025-12-12 13:26:56,729 [ 111380]   INFO - #c.i.p.w.WhatsNewShowOnStartCheckService - Checking whether to show the What's New page on startup.
2025-12-12 13:26:56,730 [ 111381]   INFO - #c.i.i.s.IdeStartupScripts - 0 startup script(s) found
2025-12-12 13:26:56,735 [ 111386]   INFO - #c.i.o.u.i.ExternalUpdateManager - update manager: Toolbox App
2025-12-12 13:26:56,744 [ 111395]   INFO - #c.i.d.PerformanceWatcherImpl - Post-startup activities under progress took 107312ms; general responsiveness: 0/45 sluggish, 45/45 very slow; EDT responsiveness: ok
```

 

Seems the keybinding for the autocompletion changed: It's now <Strg-Space> instead of <Tab>/<Enter>!!!

However, the “Suggestion as you type” is definitely broken.

2025.2.5 still worked.

“Me too”.
Some files have completely broken code inspection, giving false positives and missing actual errors.

Rolled back to 2025.2.5, which also has a nicer splash screen ;p

Has anyone tried 2025.3.1 to see if that's any better?

2025.3 and 2025.3.1 are both incredibly slow for me as well, being barely usable. I disabled all downloaded plugins and completely cleared the cache, still not usable. However, I happened to disconnect my MacBook (Pro Max M1, 64 GB) from my two external monitors so I could do some work while sitting on the sofa, and using just the MacBook's display, makes 2025.3.1 very responsive again. So I tried plugging it into just one of my two external monitors and it goes back to being barely usable. So it seems, at least for me, that external monitors are the problem. 2025.2.6 has no issue with external monitors. Is anyone else here who's having issues also using an external monitor set up?

EDIT: I found this comment about disabling the Metal renderer in the VM options, and that has made 2025.3.1 usable on my external monitors. It's still a bit laggy, but not terribly so

I can confirm that the issue is as Joe Francia has mentioned in the comment above - caused by external monitor setup.
I've downgraded to 2025.3 - a bit better, but still lagging.
Downgraded furthemore to 2025.2.6 - all fine.

FWIW, I think that the process stopping and needing to sent a SIGCONT to proceed is related to loading the Python interpreter. I opened a project that didn't have the interpeter configured and it loaded fine, as soon as I configured the interpreter the process hit a STOP and needed to be signalled to continue. (This is on Fedora 42 with Python 3.13.9)

Dima Medvedev It's ture. I'am using IntelliJ IDEA, and 2025.3.2 find not able to work(nothing find). I've downgraded, 2025.2.6 is all fine. Thanks

As the OP here. I did a fresh install on a different OS (Ubuntu 25.10) and everything works well. 

So, in case you're on a Linux based distro I suggest to apply the cleanup steps from the second post in this thread and then to do a fresh install. Afterwards be careful when enabling plugins because for example RyeCharm breaks inspections.

For me, even the older version got worse and worse. Finally, I found the issue: I started pycharm with the jetbrains toolbox installed shellscript, including a nohup and send logs to /dev/null. When opening pycharm with the gnome application launcher, everything is fine again, in old and new versions. 

I also did the “fresh install and clear everthing” approach, but it did not help in my case (besides wiping all my datagrip projects, but I won't blame anyone here, that one is on me).

Any progress on this? For me: search breaks after few minutes of running PyCharm. Clearing cache and restarting helps, but it happens again after each restart. Turns out search is something I do very often so it's not usable for me without it

Hi Yaroslav Klyuyev 

As the issues discussed here seem to have different causes/solutions and we do not have any confirmed bug on our end regarding this matter would you please try the following:

• Close the IDE

• Navigate to the project root directory and backup/delete the .idea folder (please note in UNIX based OS this is a hidden folder)

• Start the IDE

If the issue persists, would you please raise a ticket in youtrack and provide screenshot/IDE logs when this happens so we can look at this further for your specific case?

The logs can be provided from Help -> Collect Logs and diagnostics data

Kind regards,

JB

I just upgraded from 2025.2 (late last year 2025) to 2025.3.3 and it runs like glue.  What have you guys done?  Keystrokes can be seconds apart, Clicking on on windows may occur at some point – so SLOW?  This needs a fix ASAP please!

Thx
Jon
Mac OSX 26.3/MacPro/Intel/2019

I'd just like to chime in saying that I'm having the exact same problems described by Jon. Present in all versions of 2025.3.X as well as 2026.1 EAP.

2025.2.* has no issues.

Macbook Pro M3/26.2Tahoe

Hi David and Jon

As we have been unable to replicate this, we need more information and we need to look individually at each case and logs.

Therefore the best course of action would be to raise a separate case and upload logs in youtrack so we can follow up the investigations individually.

Kind regards,

As I mentioned above, I can 100% reproduce this problem every time by simply plugging in an external monitor to my MacBook. Using the MacBook alone, PyCharm 2025.3 works perfectly. I plug in a monitor (LG UltraFine 5k) and it becomes sluggish. More recent versions of 2025.3 aren't completely unusable, but are annoying enough that I've been sticking with 2025.2 when I'm using external monitors, which is most of the time

Hi everyone,

Unfortunately, we won’t be able to properly troubleshoot or provide a solution within this thread, as the issue requires a more in-depth investigation.

Please create a YouTrack issue here: https://youtrack.jetbrains.com/issues/PY

Kindly include all relevant details (logs, steps to reproduce, screenshots, sample project if applicable) to help us investigate and resolve the problem more efficiently.