improper 429 error near the end of a 5-hour window, when I still had plenty of usage left

Resolved 💬 14 comments Opened Jan 13, 2026 by dimfeld Closed Jan 13, 2026

💡 Likely answer: A maintainer (etraut-openai, contributor) responded on this thread — see the highlighted reply below.

What version of Codex is running?

custom build, slightly modified from 0.80 tag

What subscription do you have?

ChatGPT Pro

Which model were you using?

gpt-5.2-codex

What platform is your computer?

MacOS 26.2

What issue are you seeing?

I was coming up on the end of a 5-hour window, and suddenly Codex gave me 429 rate limit errors. The usage display inside Codex and also on the ChatGPT website both said I had about 61% usage remaining in my 5-hour window, which felt about right given my usage patterns for the past 5 hours.

Every time I tried a command, the "retry after" was about 60 seconds.

This issue lasted about 15 minutes, starting around 11:30 PM PST January 12, until the 5-hour window reset and things started working again.

What steps can reproduce the bug?

I don't think it will help much but here's one sample output from when it was failing.

~/Doc/cc/code3 ❯ codex exec test
OpenAI Codex v0.0.0 (research preview)
--------
workdir: /Users/dimfeld/Documents/cc/code3
model: gpt-5.2-codex
provider: openai
approval: never
sandbox: workspace-write [workdir, /tmp, $TMPDIR, /Users/dimfeld/.config/rmplan, /Users/dimfeld/.pyenv/shims, /Users/dimfeld/Library/pnpm, /Users/dimfeld/.config/chisel/logs, /Users/dimfeld/.locals/share/chisel, /Users/dimfeld/.cache, /Users/dimfeld/.semgrep] (network access enabled)
reasoning effort: medium
reasoning summaries: auto
session id: 019bb64c-b94f-7c61-98a1-61db54ac2c20
--------
user
test
mcp: rmplan starting
mcp: rmplan ready
mcp startup: ready: rmplan
2026-01-13T07:40:41.747414Z ERROR codex_api::endpoint::responses: error=http 429 Too Many Requests: Some("{\n  \"error\": {\n    \"message\": \"You've exceeded the rate limit, please slow down and try again after 60.616292 seconds.\",\n    \"type\": \"invalid_request_error\",\n    \"param\": null,\n    \"code\": \"rate_limit_exceeded\"\n  }\n}")
ERROR: exceeded retry limit, last status: 429 Too Many Requests, request id: 9bd33f31fd269bb1-HNL

What is the expected behavior?

_No response_

Additional information

_No response_

View original on GitHub ↗

14 Comments

etraut-openai contributor · 6 months ago

Please use the /feedback slash command to upload your logs and session details, then post the thread ID here.

dimfeld · 6 months ago

Just regular OpenAI hosted model. (Oh I think the comment asking about this was deleted :) )

Here's the Thread ID: 019bb644-36cd-75a3-99ba-7f037bed248a

dbxe · 6 months ago

Same here. Started about an hour ago, on a workflow that was running smoothly since end of last week.
thread ID 019bb682-7b54-7331-837b-ccab5cabc160
Stock codex-cli 0.80.0

hurxxxx · 6 months ago

exceeded retry limit, last status: 429 Too Many Requests, request id: 9bd3a3cf08d73073-ICN

Same here.

sal562 · 6 months ago

Same for me /status shows plenty of tokens left for Pro account.. but getting " exceeded retry limit, last status: 429 Too Many Requests, request id:" │
│ 5h limit: [███████████████████░] 96% left (resets 17:16) │
│ Weekly limit: [███████████████████░] 95% left (resets 11:33 on 16 Jan)

mokemokechicken · 6 months ago

thread ID 019bb6b2-8c21-7a00-91ba-60d444c0d1ae

Cardosaum · 6 months ago

thread ID 019bb7af-dfdc-72c2-a104-3be613feb1e4

lylepratt · 6 months ago

Same problem. 429 errors, plenty of usage left in my quota.

patrickvillanueva · 6 months ago

Just started getting this now

Xynonners · 6 months ago

same here. Codex is basically useless currently since it never finishes its tasks before 429'ing.

etraut-openai contributor · 6 months ago

Thanks for reporting the problem. We've addressed the underlying cause, so you should no longer see this issue.

psouza4 · 5 months ago

Thanks for reporting the problem. We've addressed the underlying cause, so you should no longer see this issue.

Just got this myself today (exceeded retry limit, last status: 429 Too Many Requests, request id: 9c5cfbb6caf1a36c-SEA). I have 81% left of my weekly quota on a Pro account plan. My prior request was 4 minutes earlier, so it'd be peculiar to hit a legitimate rate limit. GPT-5.2-Codex via Visual Studio Code v0.4.68.

Xynonners · 5 months ago

> Thanks for reporting the problem. We've addressed the underlying cause, so you should no longer see this issue. Just got this myself today (exceeded retry limit, last status: 429 Too Many Requests, request id: 9c5cfbb6caf1a36c-SEA). I have 81% left of my weekly quota on a Pro account plan. My prior request was 4 minutes earlier, so it'd be peculiar to hit a legitimate rate limit. GPT-5.2-Codex via Visual Studio Code v0.4.68.

strange, this seems to have coincided with my prompt (after thinking for 30 minutes) being shut down as " Invalid prompt: your prompt was flagged as potentially violating our usage policy. Please try again with a different prompt: https://platform.openai.com/docs/guides/reasoning#advice-on-
prompting"

what is going on in the backend?

etraut-openai contributor · 5 months ago

@psouza4, this issue was resolved. If you're seeing a similar symptom, then it's a different issue. If you see it again, use the /feedback slash command to upload your logs and session details and file a new bug.