Streaming responses are being buffered

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

I’m unable to execute prompts. cursor is unresponsive

Steps to Reproduce

It’s an issue popped up today. I have upgraded to 2.5.20 Ver for MacOS yesterday. It worked for a while and today I faced this issue.

Expected Behavior

Cursor works normally

Screenshots / Screen Recordings

Operating System

MacOS

Version Information

Version: 2.5.20
VSCode Version: 1.105.1
Commit: 511523af765daeb1fa69500ab0df5b6524424610
Date: 2026-02-19T20:41:31.942Z
Build Type: Stable
Release Track: Default
Electron: 39.4.0
Chromium: 142.0.7444.265
Node.js: 22.22.0
V8: 14.2.231.22-electron.0
OS: Darwin arm64 24.4.0

For AI issues: which model did you use?

Auto, Opus 4.6

Additional Information

I tried setting HTTP/1.1 (earlier HTTP/2) and disabling the VIT plug-in.

Does this stop you from using Cursor

Yes - Cursor is unusable

Hey, from the diagnostics screenshot it looks like Ping, Chat, and Agent are failing while everything else passes. That’s a classic sign of SSE streaming getting buffered.

You already tried HTTP/1.1 mode, nice. A few more things to try:

  1. Change DNS: try Google DNS 8.8.8.8 or Cloudflare 1.1.1.1 in System Settings > Network > DNS
  2. Test on a different network: connect via a mobile hotspot to rule out your ISP or router
  3. Try Cloudflare WARP (free VPN): this helped some users fix it
    Dishonest network error message "Result: Error: Streaming responses are being buffered by a proxy in your network environment"

By the way, are you using any VPN, proxy, or firewall? Also, what region are you in and who’s your ISP? That info can help confirm if it’s a routing issue.

Let me know what works.

#1: Already tried

#2: Already tried, but got cursor response: “Taking longer than expected” .. for a simple prompt it took longer than 35 mins and no outcome

I’m not using VPN, proxy, or firewall and I’m from India Airtel is my ISP.

Thanks for sharing the details. Since changing DNS didn’t help, and you still get “Taking longer than expected” after 35 minutes even when using a mobile hotspot, this likely isn’t just a local network issue.

A couple things to look into:

  1. Try Cloudflare WARP
    I see you tried changing DNS and using a hotspot, but have you tried Cloudflare WARP? It’s free, and a user in India with a similar issue reported that it helped. It routes traffic differently than a simple DNS change.

  2. Free plan and load
    On the free plan, requests can take much longer during peak hours since they go through a shared pool. The “taking longer than expected” behavior on a hotspot fits that. It doesn’t fully explain diagnostics failing, but it could explain waiting 35 minutes with no response.

Could you try Cloudflare WARP and let me know if anything changes? Also, what time of day (your local time) does this usually happen? Is it consistent, or does it come and go?