Extremely slow agent

Support Bug Reports
,
1

I have seen the current status page but i wonder why it is set as resolved.

Trying to work with composer 2.5 since 10am UTC+2 (Paris) and it is sooooo slooooooow.

Agent stuck at “planning next moves” indefinitely.

It is currently almost unusable you need to work this out.

Version: 3.12.17
VS Code Extension API: 1.128.0
Commit: 0fb762053c34788bb7760d5673f8a6d4c8589d50
Date: 2026-07-17T02:53:53.006Z
Layout: IDE
Build Type: Stable
Release Track: Default
Electron: 40.10.3
Chromium: 144.0.7559.236
Node.js: 24.15.0
V8: 14.4.258.32-electron.0
xterm.js: 6.1.0-beta.256
OS: Linux x64 7.0.0-28-generic

Example of chat id : 1a250b27-79b8-4232-8e32-45f0f4bd6bd8

It is a chat executing an implementation skill that takes a spec markdown to implement it and loop on some criteria until considered done. Then the agent spawns a subagent for review and loop until some criteria are met. Then spawn another subagent for security review and loop until some criteria are met.

Then the agent publish in the chat a summary for human review with a commit msg suggestion.

That is what it is done in that chat.

2

Hi there!

We detected that this may be a bug report, so we’ve moved your post to the Bug Reports category.

To help us investigate and fix this faster, could you edit your original post to include the details from the template below?

Bug Report Template - Click to expand

Where does the bug appear (feature/product)?

  • Editor, Tab & Chat (autocomplete, Composer, in-editor agent)
  • Terminal & commands
  • Models, pricing & API keys (availability, Auto/Max, BYOK/Bedrock)
  • MCP & tools
  • Cloud Agents & Automations (cursor.com/agents, scheduled/event)
  • BugBot & Code Review
  • Cursor CLI
  • Cursor Mobile
  • Remote (SSH / Dev Containers / WSL)
  • Account, billing & login
  • Something else…

Describe the Bug
A clear and concise description of what the bug is.

Steps to Reproduce
How can you reproduce this bug? We have a much better chance at fixing issues if we can reproduce them!

  1. …
  2. …
  3. …

Expected Behavior
What is meant to happen here that isn’t working correctly?

Screenshots / Screen Recordings
If applicable, attach images or videos (.jpg, .png, .gif, .mp4, .mov)

Operating System

  • Windows 10/11
  • MacOS
  • Linux

Version Information

  • For Cursor IDE: Menu → About Cursor → Copy
  • For Cursor CLI: Run agent about in your terminal
IDE:
Version: 2.xx.x
VSCode Version: 1.105.1
Commit: ......

CLI:
CLI Version 2026.01.17-d239e66

For AI issues: which model did you use?
Model name (e.g., Sonnet 4, Tab…)

For AI issues: add Request ID with privacy disabled
Request ID: f9a7046a-279b-47e5-ab48-6e8dc12daba1
For Background Agent issues, also post the ID: bc-…

Additional Information
Add any other context about the problem here.

Does this stop you from using Cursor?

  • Yes - Cursor is unusable
  • Sometimes - I can sometimes use Cursor
  • No - Cursor works, but with this issue

The more details you provide, the easier it is for us to reproduce and fix the issue. Thanks!

6

Having the same problem, really slow and stucked in “planning next moves”

The bug appears in the editor tab&chat with composer 2.5

7

Now it returns an error:

We’re having trouble connecting to the model provider. This might be temporary - please try again in a moment. Request ID: f75ba0a2-e642-4e53-9807-7188e59a4f5a

8

Hey all,

We had some capacity issues this morning. Things have been stable for the last hour or so, please give it another try!

10

It seems good now. Thanks!

12

The issue is here again with composer 2.5.

Planning next move for ever.

No issue with API models like GPT.

@Colin

14

the same, still now

15

Same here (a few days ago).

What I observed:

  • Agent UI stuck on “planning next moves”
  • Tokens still counting in Dashboard → Usage (so something was still running in the background)
  • Eventually the rest of the output showed up in the chat once the background work finished

I also noticed that after switching IDE → Dashboard → Usage → back to IDE, the UI seemed to unstick — but I’m not 100% sure that was causal. It might just have been coincidence and the chat caught up when the background work completed.

So for me it felt more like a delayed UI/status update than a fully dead agent — but take the “fix by opening Usage” part with a grain of salt.

16

Hey all!

Thanks for the reports. It’s really important to include a Request ID when reporting performance issues with a model! Without that, there isn’t much we can dig into.

17

Example right now : f4d7500c-7283-45cd-a876-3049268e249b

18

Thanks @Marlou!

That aligns with a spike in TTFT (time to first token) we saw yesterday. We have a team currently working on resolving these spikes.

For most users, these models are performing well. However, we’ve seen our p99 latency degrade intermittently over the last few days. We’re actively working on it, so please bear with us and continue reporting any particularly slow Request IDs!

19

Happens to me all the time now:

Request ID: 8bf8ee5e-691b-4e2f-9ebe-8a1f9c2e6476

Also Grok used to be FAST a week ago and now it’s as slow as Composer 2.5 (both EXTRMELY slow, minutes for simple responses)

20

Right now : 48e2e88c-4a6c-47fb-8f4f-80983e73b0f9

It is a chat where i ask the AI to implement a sequence of tasks. Started around 6pm UTC+2, it is now 8h45pm it should have finished the 5 parts i have asked for.
It is stuck on the 4th task waiting for a subagent to starts.

In the previous days i have switched from composer 2.5 to composer 2.5 fast, never had the same issue but it is sucking my plan like hell. 14 days left until limit reset and i have consumed more than 75% of composer / grok.