Using Composer 2.5 Non Fast Model - I found its too slow as compare with priviouse few days

Ideas Feedback
,
1

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

Composer 2.5 Non Fast Model I m using and it show Planning .. Planning and not making any changes i waited more than 10 min its feel like its stucked !

Steps to Reproduce

click on composer 2.5 non fast model - model is selected
now start usng agent mode - with no mcp no rule nothing just simple ui change prompt

Expected Behavior

it should work fast as it was working

Operating System

Linux

Version Information

Version: 3.11.13
VS Code Extension API: 1.125.0
Commit: 3f21b08f0b436a07be29fbfe00b304fa15553350
Date: 2026-07-10T01:45:28.254Z
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 6.8.0-134-generic

For AI issues: which model did you use?

Composer 2.5 non fast

Does this stop you from using Cursor

Yes - Cursor is unusable

4

Yeah. It started lagging yesterday afternoon, got fixed last night, but now it’s slow again.

6

Yes, it is too slow on macOS as well today.

7

every prompt is very very slow today

8

It’s still stuck on “Planning next moves”, same issue as yesterday.

9

Same here

10

me too

11

I feel bad now :face_in_clouds: :face_in_clouds: :face_in_clouds: :face_in_clouds:

12

same here. it’s still stuck on “Planning next moves” for a long time

image
image604×100 3.22 KB

13

It happened a couple of days ago on macOS as well

14

“We’re experiencing high demand for Composer 2.5 right now. Please switch to Auto, another model, or try again in a few moments.”

It’s very slow

15

Auto mode is also slow. Even with just text-based Q&A, it keeps getting stuck on “Planning next move.”

16

Hey, thanks for the report. This isn’t a bug on your side. There’s an ongoing degradation on the Composer 2.5 model provider side. Requests reach our servers, but the model is very slow to return the first token, so the agent can sit on Planning next moves for a long time. That’s why it feels like everything is stuck.

While this is ongoing, here’s a workaround:

  • Switch to Composer 2.5-fast, or another model
  • Or retry the request in a few minutes

I’ll post here once it’s stable again.

17

seems back to normal for me now

18

Hi @Dean, thanks for tuning in!

That statement makes me wonder: Aren’t you, Cursor, the model providers? Or are requests essentially being forwarded to Kimi ? If so, I assume then they are being overloaded with demand on Kimi K3 and use resources rather to satisfy demand there then on older models. And fast mode would make a difference because it routes requests to some sort of premium request tier?

Best