Same here, I ask the agent to identify the issue. It runs for two hours and get nothing. $20 are gone.
I spent $50 to solve the same issue and finally when I realized that it is Replit’s bug and asked specifically to check, I got the following reply: You’re right — I agree with your assessment. Here’s what I found:
The code is not the problem. Every time the server starts, it works perfectly — compiles pages, serves all requests with 200 status codes, no errors, no memory issues (peak 3.1GB of 8GB, zero OOM kills). The process just gets silently terminated by the platform after about 60-90 seconds.
This is a Replit platform-level behavior. I confirmed:
-
Zero OOM kills in the cgroup memory events
-
No error messages in any logs — the process just disappears
-
The exact same behavior happens regardless of the command (
npm run dev,npx next dev, with/without Turbopack, with/without memory caps) -
When run directly from the shell (not through the workflow system), the server stays alive indefinitely
So the issue is specifically with how the Replit workflow system is managing the process lifecycle right now. If there was a platform update recently, that could easily explain this — the workflow process manager may have a bug or changed behavior around idle timeouts.
1 Like
Thanks Guys! I have the same issue.. The good news is, we’re all in the same boat and it’s not our code. The bad new is, I have no idea when this will be solved. Also emailed the replit team and waiting for feedback.
They should definitely update their status page.
1 Like
Just joined Replit Pro so I could get access to this community. Glad to know I am not the only one experiencing the crashes. Like many of you - I’ve wasted a lot of money this morning trying to debug.
Any experience with getting credits added back when something like this happens? Or is it just a cost of business?
1 Like
@everyone FYI: this has been raised in an internal Slack group for the Replit team. I am sure engineers are tinkering with servers as we speak…
1 Like
I just discovered this issue… during a client demo.
1 Like
Hey all, looks like a status has been posted, you can track this to see the status change: We're currently having issues | Replit Status
1 Like
I started experiencing nonstop crashing in multiple apps at 8am CET (10 hours ago) and reported an hour or so later. I’m hoping the Replit team will be working on the immediate cause of these crashes… AND, then, also the bigger issue: why lots of smoke didn’t signal to someone (or something!) that an AI brushfire was a brewing and repeated crashes weren’t triggering alarms.
Same issue across multiple apps..
I’m having the same issue
Same issue here across multiple unrelated Node/Express apps.
Servers start successfully (port 5000 binding confirmed), DB initializes, routes return 200/304 responses, no fatal runtime errors in logs. Preview works briefly, then shows “Your app crashed.”
This feels like a runtime/container-level issue rather than app code. Also burned credits using “debug with agent” before realizing it may be systemic.
Glad to see I’m not the only one, hoping Replit can confirm what’s happening.
Hello everyone 
Apologies for the issues you faced during last nights delays.
The incident caused some apps to crash in development mode has been mitigated. We sincerely appreciate your patience while we worked through this. See here for status updates: https://status.replit.com/
If you’re still experiencing issues, please try the following steps:
• Open a new Shell tab
• Run kill 1
• Refresh your page
You can also refresh your Agent chat by clicking the chat icon in the top right corner.
If problems persist, please submit a support ticket at: https://replit.com/support and our team will be happy to help.
Thank you again for your understanding.
2 Likes
Hi,
I have a persistent issue with Replit Preview.
My server starts (port 5000), but in Preview I only see “Your app is starting” and nothing else. The app never loads.
The container keeps restarting and I constantly see:
“Restarted Start application..”
Production works perfectly.
Issue happens only in Replit Preview.
Agent debugging shows:
-
Server starts
-
Vite runs in middleware mode
-
Possibly esbuild crashes on first request
-
process.exit(1)is triggered and the process restarts -
No clear Vite error visible in logs
I already reverted all temporary debugging changes.
Problem persists.
Is anyone else experiencing Preview restart loops?
Cross posting
Hope it helps
i team,
I’m currently still experiencing critical instability with Replit Agent and workspace execution.
The Agent now frequently shows:
“Agent encountered an error while running, we are investigating the issue.”
In addition:
-
Workspace hangs on “Restarted Start application”
-
Preview/Workflow processes terminate unexpectedly
-
Shell-run works, but Agent and Workflow remain unstable
-
This happens across projects
I have an active Core subscription.
I have already contacted support and followed all suggested troubleshooting steps (kill 1, refresh, cache clear, new session, fresh projects, etc.). The issue persists.
This is now blocking active SaaS development. I’ve invested significant time and financial resources into this application (approximately $1,500 so far), and at the moment I cannot continue development due to platform instability.
Is this still part of the development-mode incident that was previously mentioned?
Can someone from the team confirm whether Agent instability is still being actively mitigated?
A clear technical update would be greatly appreciated.
1 Like
Due to recent system updates, Replit has recommended that everyone run “kill 1” in Shell if their workspace is continuing to have issues.
Open new “Shell” tab within Replit.
Type “kill 1” and hit enter
Workspace will refresh.
Hope it helps. If not, open another ticket with support.
Greeting
I do have the same issue , and I tried all the suggested work around but no luck also I submitted 3 tickets but no response from Replit, can any one help