Opus 4.6 is not responding

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

Opus 4.6 thinking is not responding.
I’ve been waiting at least 5 minutes for Claude Opus 4.6 to respond. The message “Planning next moves” is flashing. There are no network errors. I tried closing Cursor and reopening it, but it still doesn’t work. However, it is counting tokens as if it were operating.
What is going on? Is it possible that something is wrong every day?

Steps to Reproduce

Ask Opus 4.6 thinking something

Expected Behavior

Respond in normal times

Screenshots / Screen Recordings

Operating System

Windows 10/11

Version Information

Version: 2.4.36 (user setup)
VSCode Version: 1.105.1
Commit: f9919bf991f247689f9ead605b5c5a3239a2a790
Date: 2026-02-12T21:00:33.976Z
Build Type: Stable
Release Track: Early Access
Electron: 39.2.7
Chromium: 142.0.7444.235
Node.js: 22.21.1
V8: 14.2.231.21-electron.0
OS: Windows_NT x64 10.0.26200

For AI issues: which model did you use?

Opus 4.6 thinking

Does this stop you from using Cursor

No - Cursor works, but with this issue

This also happens to me when the context is full or almost full, i guess that when it enters ‘Planning next steps’ and needs to summarize the context, sometimes just fails and gets stuck there.

I usually send a new prompt and force it to get out of the queue, saying somthing like ‘Continue, you got stuck’, and 90% of the times it summarizes the context and continues (the remaining 10% gets stuck in planning next steps again)

1 Like

Thanks for the tip, tstcm.

1 Like

You’re welcome.

Anyways, it’s just a workaround, i believe there’s an underlying problem behind this.

I’ve observed that context management for opus 4.6 is buggy, usually runs out of context in the first prompt if it’s complex enough, and summarizes its context many times before giving a response (It’s even worse in 4.6 Max without max mode enabled, somehow tokens get limited to ~72k even tho it advertises 200k max context window)

I believe it got integrated too quickly without proper QA, cause its context management is far from optimal

1 Like

Having the same problem right now.

Hey, thanks for the report.

In the meantime, let’s try some diagnostics to rule out issues on your side:

  1. Disable HTTP/2: Cursor Settings > Network > HTTP Compatibility Mode > set it to HTTP/1.1. This helps a lot on some networks or VPNs.

  2. Try disabling extensions: Run cursor --disable-extensions from your command line and see if the issue still happens.

  3. Check indexing status: Look at the status bar in the bottom-right. If indexing is running, it can block the agent.

  4. Temporarily try another model: Switch to another model (for example, Claude Sonnet) to see if the issue is specific to Opus.

Also, can you share the Request ID? You can get it from the chat context menu (top-right of the chat) > Copy Request ID. This will help the team investigate.

Related thread with more context: Agent is stuck on “planning next moves”

Let me know how it goes.

Hi Dean,
I didn’t think to send you the ID, but I’ll do that next time. As for the HTTP protocol, I can tell you right away that it doesn’t solve the problem because I tried it (HTTP/1.1) while the problem was occurring. Now, however, everything seems to be working fine and responding fairly quickly.

1 Like

Hi there,

I am encountering the same issue with claude-4.6-sonnet-medium. The agent is getting stuck mid-thought for 15 min. This is request ID: 16ed35d8-cd0f-4c17-895b-55f80d947dba

Facing same issue

1 Like