Skip to content

Request timeout (504/502) - Codex local routing directly to OpenAI

Quick fix

First resolve the local routing timeout to chatgpt.com; this issue provides no confirmed additional configuration fix.

Error output
本地路由(127.0.0.1:15721)转发到上游时,直连 chatgpt.com 一直超时,请求始终无法完成。
English translation

When the local route (127.0.0.1:15721) forwards to the upstream service, direct access to chatgpt.com always times out, so requests can never complete.

Logs show the local route uses a direct connection and forwards Codex requests to https://chatgpt.com/backend-api/codex/responses. The reporter found that direct access to chatgpt.com timed out, so requests could not complete before reaching the upstream service, resulting in 504/502 errors and repeated retries.

This issue remains open, with no maintainer- or user-confirmed fix. Codex was confirmed to point to 127.0.0.1:15721/v1; OAuth login, the Plus subscription, and the token were reported as working. Changing models or restarting did not help.

  1. Confirm that this machine can connect directly to chatgpt.com; in this report, a direct-connection timeout prevents local routing from forwarding requests.

  2. If the direct connection still times out, the current issue provides no confirmed additional configuration or fix steps.

ToolCodex
VersionUnknown
PlatformsWindows

This page is distilled from 1 real issues

Is the problem with the local route's direct connection, or is extra configuration required?
This report shows that direct access to chatgpt.com timed out and logs show a direct connection; the issue provides no confirmed extra configuration solution.

This is an unofficial community wiki with no affiliation to the cc-switch authors or the project itself. Its content is compiled from the project's public GitHub issues. This site distributes no software.