Sub-agents are not using custom OpenAI base URLs

[ngochip]

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

I’ve currently reached my usage limit, and I’m using custom OpenAI base URLs while waiting for refreshes.

However, the sub-agent seems to be malfunctioning. It reports that the model is unavailable with the slow pool. This message usually occurs when I don’t enable custom OpenAI base URLs.

Note that the main agent is working fine; only the sub-agents are not functioning. Could you please tell me if this is a bug or a Cursor action to minimize the use of custom endpoints?

Steps to Reproduce

Enable a custom OpenAI URL and instruct the agent to call a sub-agent.

Expected Behavior

Sub-agents should function normally like previous versions.

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: Early Access
Electron: 39.4.0
Chromium: 142.0.7444.265
Node.js: 22.22.0
V8: 14.2.231.22-electron.0
OS: Darwin arm64 25.3.0

For AI issues: which model did you use?

Opus 4.6

Does this stop you from using Cursor

Yes - Cursor is unusable

[deanrie]

Hey, thanks for the report. Looks like this is a bug. Sub-agents aren’t inheriting your custom OpenAI Base URL settings, so they fall back to Cursor’s servers and hit the “slow pool” error because the fast requests have already been used up.

Could you please grab the Request ID from the failed sub-agent chat? Top-right chat menu > Copy Request ID. That’ll help us investigate.

[maeriyn (Martin Pan)]

Hi we’ve been running into the same problem. Are there any workarounds discovered? Is there an estimated timeline to getting this fixed?

[deanrie]

Hey, unfortunately there’s no workaround for this yet. Sub-agents don’t inherit custom API key or base URL settings, and there’s no way for the user to bypass this on their side.

I’ll update the thread when there’s a fix.

[KeentGG]

getting the same issue, i was wondering why my On Demand usage goes up when i’ve been just using my own custom API keys (OpenAI override).

So this is 99% like the cause, sub-agents are using on demand usage instead of inheriting agent model (custom model)