mirror of
https://github.com/deepseek-ai/DeepSeek-R1.git
synced 2026-11-05 00:47:42 +00:00
[GH-ISSUE #119] Server is busy for second response #98
Labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepSeek-R1#98
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @Eren-Yeaager on GitHub (Jan 28, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/119
Tried to initiate a conversation with deepseek for a project of my own . The first chat was quite clear , but from the second chat the server seems to be busy .
This problem goes away after I initiate a new chat , but it's of no use as it doesn't retain it's memory
Jam Link : https://jam.dev/c/5dabaea6-04c0-4fa0-8b9f-240da8801f9f
(Jam helps you debug quickly as it logs all the network calls )
@FullStack-webDevloper commented on GitHub (Jan 28, 2025):
the issue you mentioned is also faceby me but when i initiated a new chats the problem of server is busy persist
@Eren-Yeaager commented on GitHub (Jan 28, 2025):
Can you clear your cache and try once again ? . For me it's only for the second response
@FullStack-webDevloper commented on GitHub (Jan 28, 2025):
i have done this already the problem is still persist
@FullStack-webDevloper commented on GitHub (Jan 28, 2025):
i think the hype of deepseek result this issue
and i think the only solution for this is the scalability
@MichalZak17 commented on GitHub (Jan 28, 2025):
Correct. The servers are likely experiencing heavy load due to high traffic from users testing the model. As a temporary solution, you can access the model via API through Deepinfra or other providers. Since the model is open-source, there will likely be additional providers hosting it independently.
@ChalsLapWork commented on GitHub (Jan 28, 2025):
still busy, The server is busy. Please try again later.
@kmkamilkhel commented on GitHub (Jan 29, 2025):
still busy
@f3r33n commented on GitHub (Jan 29, 2025):
Clear cache often !
It's the only way
Btw they are getting continuously attached that's y
@swati536 commented on GitHub (Jan 29, 2025):
i am getting continue server busy response even though to start new chat also
@Abenu10 commented on GitHub (Jan 29, 2025):
same
@Shripad735 commented on GitHub (Jan 29, 2025):
Facing the same issue from last 15 Hours.
@Sahil9-coder commented on GitHub (Jan 29, 2025):
I think that when the hype decreases, the server load will also reduce. Alternatively, the company should consider increasing the server capacity.
@mihoko-da commented on GitHub (Jan 29, 2025):
no working
@Shripad735 commented on GitHub (Jan 29, 2025):
how much time it will take to solve the issue of server busy ?
@YepJin commented on GitHub (Jan 30, 2025):
when it will be back?
@sldsmkd commented on GitHub (Jan 30, 2025):
As chat length increases the amount of tokens it needs to process increases which increases the computational cost of the query. If it's on your second query then it's difficult; but I typically work around this by focussing on a single task during the context of a chat; and then ask it to provide a summary of what we discussed and findings to seed a new prompt.
@dinithaw commented on GitHub (Feb 1, 2025):
Its not a issue of model dumbo. It's just they ran out of compute power since everyone now moving towards deepseek instead of openai
@Brandyouth commented on GitHub (Feb 4, 2025):
Right! And they just HAD to charge a fee.
@utro8am commented on GitHub (Feb 20, 2025):
For those using DeepSeek Chat in the browser (not as api):
There is a free Chrome extension DeepSeek Server Busy.
It can help with "Server is busy. Please try again later." errors.
What It Does
👉 DeepSeek Server Busy Extension for auto retries.
@MuYiBo commented on GitHub (Feb 26, 2025):
Yes, I’ve encountered that too. It seems like new conversations have a higher success rate, but later ones often face server congestion. I’m using it here now, and so far, there hasn’t been any “server busy” issue, plus it’s still free. https://ai-chatbot.top/
@github-actions[bot] commented on GitHub (Apr 9, 2025):
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!
@github-actions[bot] commented on GitHub (May 9, 2025):
false
@andrewbell73 commented on GitHub (Apr 1, 2026):
Hey, I’ve seen this kind of behavior too — especially when the system is under heavy load or it starts throttling requests. There’s a common error that pops up in DeepSeek when they hit internal limits, and the best explanation I’ve found is on a troubleshooting guide that covers the “messages too frequently. try again later. deepseek” error in detail here: https://deepseeksguides.com/fix-deepseek-messages-too-frequently/
Basically, DeepSeek dynamically rate‑limits chats and API calls when the servers are busy or when too many requests are sent in quick succession, which matches what you’re describing. In your case it makes sense that the first message works and then it starts rejecting with busy/limit errors — most practical fixes revolve around slowing down requests, starting a fresh chat thread, or spacing out messages to avoid throttling.