Hitting rate limits

Resolved 💬 66 comments Opened Nov 3, 2025 by iSuslov Closed Nov 25, 2025
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

I'm aware this isn't the right place to file this issue, BUT it's now clear that a huge cut in limits is being observed. Reddit is talking about it, and your GitHub is full of similar issues. This is being systematically observed by many independent users. We have reached a point where we can look around and confirm it - we are not imagining things. And we observe silence from your end.

Is it the new standard among AI companies to serve a downgraded product and cut limits after a few month?

Thread update: ae-openai clarified that limits weren't reduced. However, a cache issue was identified which caused cache misses and huge usage spikes, with the expected fix shipped today (Nov 4).

Update 2: Today, Nov 5 usage was reset.

View original on GitHub ↗

66 Comments

github-actions[bot] contributor · 8 months ago

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

  • #6157
  • #6145
  • #6003
  • #5994
  • #5977

Powered by Codex Action

ae-openai contributor · 8 months ago

Hey,

  • We haven't reduced local limits in any way. If you're seeing lower usage, please flag it as it's a bug. On that note, we have identified one issue causing increased cache misses which are consuming limits much more quickly. Fix should be out tomorrow.
  • For cloud (aka web), we are working to enable those limits to go much further. We just shipped our first improvement tonight: Failed cloud tasks will no longer count against limits. (We should have caught this earlier.) Many more improvements coming.
lunamidori5 · 8 months ago
* For cloud (aka web), we are working to enable those limits to go much further. We just shipped our first improvement tonight: Failed cloud tasks will no longer count against limits. (We should have caught this earlier.) Many more improvements coming.

Thank you! On top of that I hope those poorly formatted messages that crash the full cloud website also do not count... Thank y'all for your hard work!

iSuslov · 8 months ago

Hey @ae-openai, this is my particular case:

<img width="2172" height="1956" alt="Image" src="https://github.com/user-attachments/assets/3b057c0e-c078-4705-8b09-19ad80d3b4d4" />

It says limits will reset on Nov 9th, thus it started on Nov 2th and by using it for one day (two 5 hour windows) I've got 80% of my weekly limit reached. You guys should announce this caching issue publicly and reset weekly limits of those who was affected.

techjoec · 8 months ago

Reposting some data here as https://github.com/openai/codex/issues/6008#issuecomment-3483530072 (@etraut-openai ) noted this thread is an "update". Issue I put data into on Saturday post "fix": https://github.com/openai/codex/issues/6008

Summary

Plan: PRO
Duration to weekly limit hit: Under 24 hours!
Usage "week" started: 11/2/2025 12:10:00 AM UTC.

My week

I ran my numbers for CLI and cloud from the start of my "week" til I was prevented from using the service and this is what I came up with. Might be off a slight hair but likely very close to actuals..

Cloud Totals
CloudTasks: 109
Tokens: 3891767
Total Task Hours: 20.25

Cloud Task Averages
Tokens per task: 35704
Minutes per task: 11

CLI Session Totals
InputCachedTokens - 352834688
InputTokens - 375819446
OutputTokens - 2073269
ReasoningTokens - 1042304

Five-hour usage window sample

As for 5 hour usage...The below activity (and the CLI is only end of session stats, and some were likely started before that current 5-hour window reset so likely a bit _inflated_!

Cloud Totals
CloudTasks: 29
Tokens: 1075448
Total Task Hours: 5.38

Cloud Task Averages
Tokens per task: 37084
Minutes per task: 11

CLI Session Totals
InputCachedTokens - 90468224
InputTokens - 96415999
OutputTokens - 653123
ReasoningTokens - 351552

Serena MCP is the only extra tool I use, and Codex CLI makes sure to forget about it for 99% of the session anyway.

Overall 5 hour span of usage
| timestamp | source | primary | secondary | InputTokens | InputCachedTokens | OutputTokens | ReasoningTokens | assistant_turn_tokens | duration_minutes | diff_adds | diff_removes |
| ------------------------ | ------ | ------- | --------- | ----------- | ----------------- | ------------ | --------------- | --------------------- | ---------------- | --------- | ------------ |
| 2025-11-02T00:02:12.57Z | CLI | 0 | 11 | 10301592 | 9392512 | 82209 | 33088 | | | | |
| 2025-11-02T00:08:10.199Z | CLI | 1 | 11 | 165130 | 140160 | 5538 | 4224 | | | | |
| 2025-11-02T00:09:06.164Z | CLI | 1 | 12 | 216218 | 167552 | 8636 | 6720 | | | | |
| 2025-11-02T00:09:55.806Z | CLI | 2 | 12 | 11143894 | 10085888 | 94728 | 39040 | | | | |
| 2025-11-02T00:17:29.111Z | CLI | 3 | 12 | 849279 | 724608 | 11920 | 6656 | | | | |
| 2025-11-02T00:19:45.013Z | CLI | 4 | 12 | 16552345 | 14958976 | 108591 | 44096 | | | | |
| 2025-11-02T00:20:26.739Z | CLI | 5 | 12 | 1574876 | 1427200 | 20049 | 8384 | | | | |
| 2025-11-02T00:20:28Z | CLOUD | | | | | | | 26283 | 7 | 279 | 92 |
| 2025-11-02T00:21:33.769Z | CLI | 5 | 13 | 18241249 | 16472448 | 112098 | 45696 | | | | |
| 2025-11-02T00:24:35Z | CLOUD | | | | | | | 32075 | 9 | 542 | 542 |
| 2025-11-02T00:28:16.486Z | CLI | 6 | 13 | 22166415 | 20332672 | 123165 | 50304 | | | | |
| 2025-11-02T00:30:21Z | CLOUD | | | | | | | 56573 | 18 | 367 | 184 |
| 2025-11-02T00:31:04.325Z | CLI | 7 | 13 | 23844089 | 21805952 | 128001 | 52480 | | | | |
| 2025-11-02T00:36:40Z | CLOUD | | | | | | | 46140 | 10 | 1062 | 6 |
| 2025-11-02T00:50:15Z | CLOUD | | | | | | | 126616 | 26 | 927 | 31 |
| 2025-11-02T01:03:04.502Z | CLI | 8 | 13 | 4172315 | 3762432 | 60647 | 35264 | | | | |
| 2025-11-02T01:09:16.943Z | CLI | 8 | 14 | 24411605 | 22370176 | 129541 | 53184 | | | | |
| 2025-11-02T01:17:35.241Z | CLI | 11 | 3 | 7467843 | 6985216 | 87521 | 50496 | | | | |
| 2025-11-02T01:20:19Z | CLOUD | | | | | | | 59472 | 14 | 837 | 108 |
| 2025-11-02T01:21:38.957Z | CLI | 12 | 3 | 8586939 | 8084608 | 95488 | 54080 | | | | |
| 2025-11-02T01:30:50.153Z | CLI | 12 | 4 | 8930397 | 8315776 | 103363 | 59008 | | | | |
| 2025-11-02T01:35:16Z | CLOUD | | | | | | | 47974 | 12 | 1226 | 20 |
| 2025-11-02T01:36:18.228Z | CLI | 17 | 5 | 10658080 | 9909120 | 116252 | 65792 | | | | |
| 2025-11-02T01:37:31.099Z | CLI | 18 | 5 | 11331541 | 10577664 | 121280 | 68224 | | | | |
| 2025-11-02T01:38:56.01Z | CLI | 18 | 6 | 11727535 | 10959488 | 125171 | 70848 | | | | |
| 2025-11-02T01:47:04.227Z | CLI | 19 | 6 | 13697296 | 12771584 | 140823 | 77632 | | | | |
| 2025-11-02T01:48:41.797Z | CLI | 22 | 7 | 13697296 | 12771584 | 140823 | 77632 | | | | |
| 2025-11-02T01:50:24Z | CLOUD | | | | | | | 41994 | 9 | 907 | 74 |
| 2025-11-02T01:54:10Z | CLOUD | | | | | | | 10656 | 10 | 68 | 5 |
| 2025-11-02T01:58:21.991Z | CLI | 23 | 7 | 15007732 | 14055296 | 164822 | 99520 | | | | |
| 2025-11-02T02:00:25Z | CLOUD | | | | | | | 89546 | 22 | 1482 | 27 |
| 2025-11-02T02:04:02.047Z | CLI | 28 | 9 | 15305937 | 14351488 | 169008 | 103168 | | | | |
| 2025-11-02T02:06:15Z | CLOUD | | | | | | | 18869 | 9 | 183 | 14 |
| 2025-11-02T02:14:29.01Z | CLI | 29 | 9 | 57367 | 37888 | 754 | 512 | | | | |
| 2025-11-02T02:16:04.433Z | CLI | 32 | 9 | 18085223 | 17096064 | 185383 | 110272 | | | | |
| 2025-11-02T02:16:26.43Z | CLI | 32 | 10 | 687054 | 573440 | 6163 | 3968 | | | | |
| 2025-11-02T02:18:49Z | CLOUD | | | | | | | 12103 | 9 | 103 | 22 |
| 2025-11-02T02:20:37.583Z | CLI | 33 | 10 | 2896820 | 2571136 | 16893 | 10176 | | | | |
| 2025-11-02T02:25:22Z | CLOUD | | | | | | | 71519 | 20 | 901 | 43 |
| 2025-11-02T02:30:25.715Z | CLI | 45 | 14 | 19385592 | 18358272 | 192675 | 113984 | | | | |
| 2025-11-02T02:30:59Z | CLOUD | | | | | | | 41871 | 10 | 254 | 5 |
| 2025-11-02T02:31:05.065Z | CLI | 46 | 14 | 20036967 | 19006592 | 194727 | 114112 | | | | |
| 2025-11-02T02:43:06Z | CLOUD | | | | | | | 29800 | 13 | 148 | 6 |
| 2025-11-02T02:43:11.204Z | CLI | 49 | 15 | 20701246 | 19655296 | 199937 | 116672 | | | | |
| 2025-11-02T02:46:24.807Z | CLI | 56 | 17 | 21391911 | 20331264 | 204662 | 119232 | | | | |
| 2025-11-02T02:50:19Z | CLOUD | | | | | | | 42321 | 10 | 566 | 16 |
| 2025-11-02T02:52:48.35Z | CLI | 57 | 17 | 22814134 | 21563136 | 210774 | 122112 | | | | |
| 2025-11-02T03:00:35Z | CLOUD | | | | | | | 15494 | 7 | 81 | 7 |
| 2025-11-02T03:00:50.215Z | CLI | 63 | 19 | 24473416 | 23202048 | 220054 | 125696 | | | | |
| 2025-11-02T03:01:48.331Z | CLI | 66 | 20 | 24473416 | 23202048 | 220054 | 125696 | | | | |
| 2025-11-02T03:05:28Z | CLOUD | | | | | | | 36797 | 10 | 291 | 39 |
| 2025-11-02T03:11:32.052Z | CLI | 69 | 21 | 25981713 | 24701184 | 226658 | 128832 | | | | |
| 2025-11-02T03:12:47Z | CLOUD | | | | | | | 17178 | 9 | 267 | 10 |
| 2025-11-02T03:20:17Z | CLOUD | | | | | | | 34360 | 8 | 235 | 10 |
| 2025-11-02T03:24:51Z | CLOUD | | | | | | | 35604 | 18 | 106 | 25 |
| 2025-11-02T03:29:46.233Z | CLI | 78 | 23 | 26753057 | 25467008 | 228396 | 129216 | | | | |
| 2025-11-02T03:30:16Z | CLOUD | | | | | | | 40582 | 11 | 5928 | 12 |
| 2025-11-02T03:38:08Z | CLOUD | | | | | | | 16313 | 6 | 258 | 0 |
| 2025-11-02T03:43:13Z | CLOUD | | | | | | | 36926 | 16 | 198 | 24 |
| 2025-11-02T03:45:11Z | CLOUD | | | | | | | 21209 | 5 | 473 | 23 |
| 2025-11-02T03:48:55.153Z | CLI | 95 | 29 | 9546 | 0 | 71 | 64 | | | | |
| 2025-11-02T03:50:19Z | CLOUD | | | | | | | 28764 | 8 | 542 | 34 |
| 2025-11-02T03:52:47Z | CLOUD | | | | | | | 0 | | | |
| 2025-11-02T03:54:16Z | CLOUD | | | | | | | 35176 | 16 | 209 | 0 |
| 2025-11-02T03:54:56Z | CLOUD | | | | | | | 3233 | 1 | 7 | 1 |
| 2025-11-02T04:10:04.832Z | CLI | 100 | 30 | | | | | | | | |

Daily usage from OpenAI directly

Unsure if https://chatgpt.com/backend-api/wham/usage/credit-usage-events reports "date" in UTC or not...But if it does, then the first day can be fully ignored. But it doesn't even add up to 100%...Confusing.

{
"date": "2025-11-01",
"product_surface_usage_values": {
"cli": 4.734244660872603,
"web": 19.14151781164927,
"github_turn": 0.06114338122898196,
"github_code_review": 0.7460948304726966
}
}

{
"date": "2025-11-02",
"product_surface_usage_values": {
"cli": 5.380617548150413,
"web": 39.7206329795752,
}
},

Soooooooooooooo

@etraut-openai This is "working as intended"? Please advise ASAP.

The site says:
Pro Usage: Usage limits apply across both local and cloud tasks. Average users can send 300-1,500 local messages every 5 hours or 50-400 cloud tasks every 5 hours, with a shared weekly limit.

That is _really_ not the case here it seems.

msbealo · 8 months ago

I'm not a fancy CLI user, but I have been using the web interface for some projects. However, on Thursday, the 30th, I was hit by this issue. I was suddenly rate-limited, significantly below my previous usage. I struggled on for a few days using a local environment in VS Code, but it didn't help much. My 5-hour limit was being reached after 2-3 prompts. Now my weekly limit has been reached, and I have to wait 5 more days!

In the meantime, I've been investigating alternatives, but this has significantly affected my progress.

<img width="1271" height="783" alt="Image" src="https://github.com/user-attachments/assets/8e8c0521-1990-4ac5-80fc-854e407466a5" />

pontes-a7 · 8 months ago

any update? @ae-openai

techjoec · 8 months ago

The only update I've seen is that they will stop charging us for when Cloud fails. That should be a no-brainer and should have already been in place before billing us, but that aside there's something seriously wrong OR the dozens/hundreds of us complaining (see OpenAI's discussion boards, reddit.com/r/codex/, here, etc) are either stupid or crazy (tell us which, please). I've spent (wasted) hours providing numbers and many others have too. OAI seems to feel this is working as intended apparently.

These comments here are frustrating: https://old.reddit.com/user/embirico especially "web tasks are longer". Uhhh..I average 11 minutes overall. ~100 cloud tasks and ~22mil non-cached tokens from CLI ate 100% of my Pro weekly limit!?!?

As for CLI usage, this is a recent dump I did of the token usage. I haven't really changed my habits at all, in fact have done things to try to streamline, but apply_patch and sed taking up such a gigantic share of the token load, and often on the same damn files over and over, surely is not helping. It feels like after making such a gigantic usage-impacting change with sudden jerk-of-the-wheel motions that slow-rolling tiny "improvements" now really isn't helping one bit here.

| Function call type | Calls | Total tokens | Share of total | Avg tokens per call |
| -------------------------------------------------- | --------: | --------------: | -------------: | ------------------: |
| shell | 1,315 | 302,839,198 | 88.86 % | ≈230,296 |
| mcp__serena__write_memory | 26 | 8,277,500 | 2.43 % | 318,365 |
| update_plan | 60 | 7,582,367 | 2.22 % | 126,373 |
| mcp__serena__search_for_pattern | 51 | 5,612,423 | 1.65 % | 110,048 |
| mcp__serena__read_memory | 42 | 4,086,937 | 1.20 % | 97,308 |
| mcp__serena__think_about_whether_you_are_done | 11 | 2,875,806 | 0.84 % | 261,437 |
| mcp__serena__find_symbol | 14 | 1,486,131 | 0.44 % | 106,152 |
| mcp__serena__summarize_changes | 3 | 1,046,374 | 0.31 % | 348,791 |
| mcp__serena__think_about_task_adherence | 4 | 939,356 | 0.28 % | 234,839 |
| mcp__serena__list_dir | 16 | 894,987 | 0.26 % | 55,937 |
| mcp__serena__get_current_config | 4 | 868,927 | 0.25 % | 217,232 |
| mcp__serena__think_about_collected_information | 6 | 749,603 | 0.22 % | 124,934 |
| mcp__serena__check_onboarding_performed | 11 | 708,449 | 0.21 % | 64,404 |
| mcp__serena__find_file | 8 | 564,971 | 0.17 % | 70,621 |
| mcp__serena__initial_instructions | 5 | 530,192 | 0.16 % | 106,038 |
| mcp__serena__list_memories | 3 | 438,866 | 0.13 % | 146,289 |
| mcp__serena__activate_project | 13 | 382,333 | 0.11 % | 29,410 |
| mcp__serena__switch_modes | 8 | 381,639 | 0.11 % | 47,705 |
| mcp__serena__get_symbols_overview | 3 | 294,828 | 0.09 % | 98,276 |
| read_mcp_resource | 3 | 80,636 | 0.02 % | 26,879 |
| list_mcp_resources | 2 | 72,177 | 0.02 % | 36,089 |
| list_mcp_resource_templates | 1 | 50,198 | 0.01 % | 50,198 |
| mcp__serena__onboarding | 1 | 39,656 | 0.01 % | 39,656 |
| Other / Unassigned | — | 12,802 | 0.00 % | — |

| Command | Calls | Total tokens | Share of shell tokens | Avg tokens/call |
| --------------------------------- | ----: | -------------: | --------------------: | --------------: |
| apply_patch | 250 | 78,625,822 | 26.00 % | ≈314,503 |
| sed | 239 | 43,584,953 | 14.41 % | 182,363 |
| lua | 114 | 38,096,266 | 12.60 % | 334,177 |
| rg (ripgrep) | 144 | 24,911,137 | 8.24 % | 172,994 |
| ****** | 95 | 24,834,789 | 8.21 % | 261,418 |
|
python3 | 40 | 15,988,559 | 5.29 % | 399,713 |
|
git | 57 | 9,117,436 | 3.02 % | 159,955 |
|
ls | 69 | 8,844,375 | 2.92 % | 128,179 |
|
nl | 55 | 8,468,044 | 2.80 % | 153,964 |
|
pnpm | 29 | 6,715,358 | 2.22 % | 231,564 |
|
luacheck | 23 | 6,227,061 | 2.06 % | 270,741 |
|
applypatch | 23 | 5,437,445 | 1.80 % | 236,410 |
|
python | 22 | 5,112,512 | 1.69 % | 232,386 |
|
cat | 30 | 4,911,715 | 1.62 % | 163,723 |
|
jq | 18 | 2,287,776 | 0.76 % | 127,098 |
|
make_package.sh | 4 | 2,166,519 | 0.72 % | 541,629 |
|
set | 10 | 1,501,474 | 0.50 % | 150,147 |
|
gh | 3 | 1,455,285 | 0.48 % | 485,095 |
|
find | 11 | 1,418,196 | 0.47 % | 128,926 |
|
rm | 6 | 1,384,068 | 0.46 % | 230,678 |
|
mkdir | 5 | 1,351,405 | 0.45 % | 270,281 |
|
lynx | 7 | 1,203,903 | 0.40 % | 171,986 |
|
chmod | 3 | 940,198 | 0.31 % | 313,399 |
|
luac | 5 | 915,651 | 0.30 % | 183,130 |
|
bundle=/tmp/task_bundle.json; | 8 | 855,772 | 0.28 % | 106,971 |
|
echo | 5 | 794,555 | 0.26 % | 158,911 |
|
head | 5 | 769,247 | 0.25 % | 153,849 |
|
wc | 4 | 589,987 | 0.20 % | 147,496 |
|
unzip | 5 | 585,306 | 0.19 % | 117,061 |
|
printf | 3 | 519,334 | 0.17 % | 173,111 |
|
grep | 5 | 465,364 | 0.15 % | 93,072 |
|
awk | 2 | 387,552 | 0.13 % | 193,776 |
|
for | 2 | 379,284 | 0.13 % | 189,642 |
|
rmdir | 1 | 372,822 | 0.12 % | 372,822 |
|
timeout_ms | 1 | 366,519 | 0.12 % | 366,519 |
|
timeout | 2 | 181,203 | 0.06 % | 90,601 |
|
bash | 1 | 130,110 | 0.04 % | 130,110 |
|
tmpdir=$(mktemp) | 1 | 83,302 | 0.03 % | 83,302 |
|
printenv | 1 | 61,054 | 0.02 % | 61,054 |
|
UA="Mozilla/5.0 | 1 | 60,877 | 0.02 % | 60,877 |
|
TOKEN=$(jq | 1 | 56,533 | 0.02 % | 56,533 |
|
curl | 2 | 47,648 | 0.02 % | 23,824 |
|
which** | 1 | 28,485 | 0.01 % | 28,485 |

I hate being a whiner, but at $200/month having my entire weekly quota blown in under 24 hours is absolutely bonkers and not aligned with the stated plan limits from what I can tell. And my usage is hobby strictly, OAI is banned and CIDR blocked at work for Reasons(tm). I feel for the folks getting jammed with this that rely on it for their day jobs!

<img width="1258" height="749" alt="Image" src="https://github.com/user-attachments/assets/e306f779-88a4-42d1-8a20-d8eca9a9bf9c" />

Also this one is gold right here: https://github.com/openai/codex/issues/6198 Poor guy had codex change one line and lost 8% of his Plus budget. Yeah, tell us how that is "working correctly" please??

techjoec · 8 months ago

And to add, the messaging from OAI is borderline sloppy. Saw a few tweets today now that "in the next two weeks" things might improve. Then nuggets like this:

<img width="1133" height="790" alt="Image" src="https://github.com/user-attachments/assets/ad1130e0-7f9b-4919-99cd-91cddc642dec" />

So which is it, team? Are us users stupid/crazy, or is the platform broken??

techjoec · 8 months ago

Ah well. Canceled my Pro. Guess that's all I can do. Best of luck to all and I hope they fix things. :(

The interesting part of it is that I tried to get the "AI Support" to open a ticket for me before doing this, but it insisted that my usage is advanced and suggested I pay more for credits. After a half dozen prompts I said "Ok cancel my Pro" and without any hesitation it fully canceled my plan. No confirmations or anything...ha! I was going to keep Plus for a month, but the kick in the butt was clear!

You said: Ok cancel my Pro The assistant said: You have been granted a prorated refund based on your usage. Your active subscription has been cancelled. The refund will show up in your bank statement within 10 days.
aitikcho · 8 months ago

It was a good cloud for coding, however ... today I have tried 3 tasks, all of them failed :
By the time of these 3 unsuccessful tasks I am 80% of my 5 hour limit and 25% of my 5 day limit. I used to code with this tool and now it seems I will have to use another tool. Hope I get good suggestions.

I’m sorry, but I’m not able to complete this request.

why?

Copy
I’m not sure why the earlier response said the request couldn’t be completed—there’s nothing in the project setup that prevents working on it. If you’d still like to proceed, let me know and I’ll be happy to help implement the project-specific data handling and dropdown.

msbealo · 8 months ago

It looks like my usage was reset this evening. I don't know if this has been applied to others, or if it's related to a support ticket I put in a couple of days ago, or what. While I appreciate the reset, I also notice that ONE prompt has used up 17% of my 5-hour usage. It's nice to be back. I certainly haven't got Kilo Code working as well as Codex does on my project, but I hope that this isn't just a temporary fix.

<img width="1308" height="844" alt="Image" src="https://github.com/user-attachments/assets/0cae7381-72c0-4773-a208-5dd243550601" />

techjoec · 8 months ago

https://old.reddit.com/r/codex/comments/1oplu4l/200_in_free_credits_for_cloud_users/

This is some BS, if you downgraded because of it you don't get the $200 in credits. Ah well, posting here for anyone still eligible.

It’s been an eventful week. To thank you for your patience, we’ve granted $200 in free credits to Plus and Pro users who used cloud tasks in the past month, valid until Nov 20. You see them on your usage page. In the coming days, very excited to share more updates that will help everyone get more usage of Codex on cloud and everywhere else
defendicon · 8 months ago

Mine reset also and got 5000 credit refund. and only 1 prompt with 4 consecutive container task use almost 5 hours limit. Its not useable any more.

<img width="1258" height="928" alt="Image" src="https://github.com/user-attachments/assets/8c461bd7-66d4-4102-b309-4f247bb156b2" />

etraut-openai contributor · 8 months ago

Here's another update from the codex team.

We've been working to fix a number of issues with tracking cloud task usage. Some of these improvements have been rolled out already, and several more are still in the works. We'll post more updates here in the upcoming days.

To thank you for your patience, we’ve granted $200 in free credits to Plus and Pro users who used cloud tasks in the past month. These are valid until Nov 20.

techjoec · 8 months ago

What about those of us that canceled our sub after this fiasco? Will we get the $200 credit too?

rymatoi · 8 months ago
What about those of us that canceled our sub after this fiasco? Will we get the $200 credit too?

didnt canceled my sub, my usage was reset, but I didnt get any of the credits ..

prabhatpushp1997 · 8 months ago

I just upgraded to plus to use codex and I reached my 5 hours limit with a single message. Is this expected. I am frustrated, claude-code works much better than this. Attaching screenshots. Also I noticed openai does not have a live customer support. only ai support and general FAQ.😭

<img width="934" height="704" alt="Image" src="https://github.com/user-attachments/assets/f2e1448b-5b5b-4b3e-8f8e-2561c0261761" />

<img width="209" height="44" alt="Image" src="https://github.com/user-attachments/assets/56015619-45df-4ca6-b18c-8b8c4ab0ed7d" />

<img width="1439" height="767" alt="Image" src="https://github.com/user-attachments/assets/5976ddb5-fba8-4390-9ab5-0cd802d3f55e" />

Dylmoco · 8 months ago

usage limit still being rinsed with barely any usage since claimed fixes. Also cant see limit usage as its blanked out with "we couldn't load your usage" very annoying. Using codex via coud.

<img width="1076" height="1015" alt="Image" src="https://github.com/user-attachments/assets/a5ee4d1f-2ff0-45dd-b7a9-4c4e29236c4e" />

techjoec · 8 months ago

@etraut-openai @tibo-openai

I asked the AI support about the credit in the same thread it canceled my sub in, and it actually 'opened' a ticket. This is the response I got. Can I get a response here please?

The "Official" stance from OpenAI support here is:

However, there is no official documentation confirming an automatic $200 Codex credit or compensation program associated with this issue. Recent communications about Codex credits refer to the new pay-as-you-go system for Plus and Pro users, which allows users to purchase credits for continued usage beyond plan limits. These credits are not automatically granted.

---------------
email with support below
---------------

From: Support Date: Thu, Nov 6, 2025 at 7:17 PM --- Hello, Thank you for reaching out to OpenAI Support. We understand you’d like confirmation regarding your refund and the status of your ChatGPT Pro subscription. We know it’s important to have clarity on your billing activity, and we’re here to assist. Our ChatGPT subscription fees are generally non-refundable, as outlined in our Terms of Use. However, we understand that purchases sometimes don't work out, and we're sorry to know you wish to cancel and receive a refund. Upon checking your account, we can confirm that you no longer have an active ChatGPT subscription. After further review, we can verify that your ChatGPT Pro refund of $25.81 has already been processed as a prorated refund, since your request was made 27 days after your most recent subscription charge. Please note that refunds may take up to 10 business days to appear on your statement, depending on your bank or payment provider. If you have any other questions or need additional assistance, please feel free to reach out. We’re here to help. Best, Marvin Kyser OpenAI Support

---

From: [REDACTED] Date: Thu, Nov 6, 2025 at 10:23 PM --- I am seeking the $200 Codex web credit. I reported the issue on GitHub in extreme detail and suspended my subscription due to the fact I was at quota after about an hour for the entire week suddenly. The Codex project team has admitted fault and granted $200 to Plus and Pro users who used cloud tasks in the past month. I definitely have. I am techjoec in this issue on the official repo and per the Codex team: 🔗 https://github.com/openai/codex/issues/6172#issuecomment-3499330871

---

From: Support Date: Fri, Nov 7, 2025 at 5:11 PM --- Hello, Thank you for reaching back to OpenAI Support. We understand you’re referring to the $200 Codex web credit mentioned in connection with recent Codex updates. We know how important it is to have clarity about any credits or adjustments, and we appreciate your patience while we reviewed this. After checking your account, we can confirm that a $200 refund was issued on November 5, 2025 to your original payment method, and your ChatGPT Pro subscription was canceled the same day. However, there is no official documentation confirming an automatic $200 Codex credit or compensation program associated with this issue. Recent communications about Codex credits refer to the new pay-as-you-go system for Plus and Pro users, which allows users to purchase credits for continued usage beyond plan limits. These credits are not automatically granted. If you have any other questions or need further assistance, please feel free to reach out. We’re here to help.

---

From: [REDACTED] Date: Fri, Nov 7, 2025 at 7:09 PM --- 1. Are you a person or AI? 2. The refund was $200? Can you read the system again? 3. I am sharing this email with the public and the OpenAI Codex staff that would probably disagree about the credit situation. 4. I won't even comment on this obviously AI-generated line beyond this. 5. Escalate this as a real ticket and respond immediately with a tracking/ticket number please. Best, Joe
oldandangry · 8 months ago

The rate limits are absolutely insane. On the plus plan the weekly limit is 15-20% higher than the 5 hour limit, and the 5 hour limit is pathetic. I'm moving on to Claude once I've used up the credits.

techjoec · 8 months ago
Support Sat, Nov 8, 4:31 PM (11 hours ago) to me Hello, Thank you for reaching back to OpenAI Support. We understand you’d like confirmation about the $200 refund and additional details regarding the Codex credit. We know how important it is to have accurate information about your billing and account activity, and we’re here to assist. After reviewing your account again, we can confirm that a $200 prorated refund was successfully issued to your original payment method, and your ChatGPT Pro subscription was canceled the same day. Please note that refunds may take up to 10 business days to appear on your statement, depending on your bank or payment provider. At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records. We appreciate you sharing the GitHub link for context. While we’re unable to confirm any program granting automatic Codex credits, your refund has been fully processed, and no further billing adjustments appear pending. If you have any other questions or need assistance, please feel free to reach out. We’re here to help.

WTF?

1) my refund is only ~$25.
2) Apparently support STILL, as of November 8th, doesn't know a thing about the Codex credits? "we’re unable to confirm any program granting automatic Codex credits,"
3) "At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records." DUH that's why I opened a ticket. Still won't give me a ticket number either.

@ae-openai @etraut-openai @tibo-openai @bolinfest @gpeal Anyone???

rymatoi · 8 months ago
> Support > Sat, Nov 8, 4:31 PM (11 hours ago) > to me > Hello, > Thank you for reaching back to OpenAI Support. > We understand you’d like confirmation about the $200 refund and additional details regarding the Codex credit. We know how important it is to have accurate information about your billing and account activity, and we’re here to assist. > After reviewing your account again, we can confirm that a $200 prorated refund was successfully issued to your original payment method, and your ChatGPT Pro subscription was canceled the same day. Please note that refunds may take up to 10 business days to appear on your statement, depending on your bank or payment provider. > At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records. We appreciate you sharing the GitHub link for context. While we’re unable to confirm any program granting automatic Codex credits, your refund has been fully processed, and no further billing adjustments appear pending. > If you have any other questions or need assistance, please feel free to reach out. We’re here to help. WTF? 1. my refund is only ~$25. 2. Apparently support STILL, as of November 8th, doesn't know a thing about the Codex credits? "we’re unable to confirm any program granting automatic Codex credits," 3. "At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records." DUH that's why I opened a ticket. Still won't give me a ticket number either. @ae-openai @etraut-openai @tibo-openai @bolinfest @gpeal Anyone???

they ignore the problem closing issues with complains to make it look like it is not the global problem, they wait till 20 nov to officially announce these limitations or smth
they redirect every closed issue to this thread and then just abandon it for 2 weeks already, a problem that is not even a problem, noone wants to solve it because it's not a problem, they planned to limitate usage to earn more money, they wouldnt have integrated 'credits' without those limitations just because noone would use them as noone never hit usage limit before (i didnt even know it existed till Nov)
they have huge problems with it, they waste more than earn despite official say
I understand it that way
and the saddest part is that they try to hide it

KAMKEEL · 8 months ago

Yeah, I've experienced the rate limits being hit extremely inconsistently. It's become very unusable at this point. Likely going to have to find an alternative because I really enjoyed the Codex-Web version for queuing off tasks when I was unavailable from my PC.

Dylmoco · 8 months ago

Guys they have completely Fu***d the rate limits, the past 1-2 weeks it has been pretty much useless, and 2-3 tasks my 5 hours gone and my weekly is rinsed. I don't know what they have done, but what I do know, its unusable, and ill be moving to a different platform

ae-openai contributor · 8 months ago

Hey folks, rate limits for Codex CLI and IDE extension are in a good spot at the moment. We actually increased them significantly for Plus users last week.

We know cloud limits are still low. We've increased them slightly, but are hoping to do much more there. We've also seen some reports of folks not getting credits, and we're looking into them. The team is on it!

I've also just reached out to the support to let them know of the situation.

pfahlr · 8 months ago

I haven't checked to see how quickly limits are used since Nov 6, but the usage for nov 5 and 6 (which is clearly less than on oct 21/22/23 - and oct 30) used up over 77% of my monthly quota as well as nearly 5000 purchased credits.

Not sure if I should have expected to see the remaining usage restored... not that it matters much at this point if it stops using so much per operation... will update once I've run a few codex operations.

<img width="1147" height="717" alt="Image" src="https://github.com/user-attachments/assets/468f151a-9de1-49fe-a41b-f80f427d61a6" />

pfahlr · 8 months ago

still doesn't look good

<img width="1168" height="438" alt="Image" src="https://github.com/user-attachments/assets/14d34bc0-a42b-4c0a-bfe6-f087c5942549" />

ran 1 more operation and another running now.

pfahlr · 8 months ago

<img width="1139" height="461" alt="Image" src="https://github.com/user-attachments/assets/c47deb4a-d151-4006-8c96-f1828ee79b37" />

and here's after 1 more... running codex CLI in a local enviornment.

pfahlr · 8 months ago

after 1 more codex CLI operation in local environment

<img width="1137" height="477" alt="Image" src="https://github.com/user-attachments/assets/74d8b498-0a59-4ffe-b577-518732f7b041" />

you can see from the time between posts, this is running a single thread of codex CLI on my local machine.

pfahlr · 8 months ago

<img width="1124" height="398" alt="Image" src="https://github.com/user-attachments/assets/00d719bb-fa7d-409a-b4e8-1f7fe4c4110d" />

and another... same thing

pfahlr · 8 months ago

<img width="1139" height="403" alt="Image" src="https://github.com/user-attachments/assets/6a6fd3a0-dbee-49e4-92b7-1a45ec89d7d7" />

after 1 more

pfahlr · 8 months ago

<img width="1126" height="391" alt="Image" src="https://github.com/user-attachments/assets/743bb7f8-1871-4c45-b199-595df5567262" />

going... going...

pfahlr · 8 months ago

<img width="1147" height="394" alt="Image" src="https://github.com/user-attachments/assets/eb337431-36d0-4f7a-8903-950e9f30d940" />
and gone!

this is giving me so much less than it was a few weeks ago...

I was unable to use up my limits until now... now I can do it in a couple sessions.

pfahlr · 8 months ago
The rate limits are absolutely insane. On the plus plan the weekly limit is 15-20% higher than the 5 hour limit, and the 5 hour limit is pathetic. I'm moving on to Claude once I've used up the credits.

https://code.claude.com/docs/en/overview

hmm... they have the same kinda thing, huh?

KAMKEEL · 8 months ago

<img width="1280" height="805" alt="Image" src="https://github.com/user-attachments/assets/abfbeee7-65da-4784-8029-5a42b40977df" />

I'm not sure; the limit hitting has been quite brutal recently. I've already hit my limit after 6 tasks total today and as you can see on Nov 13, it's barely anything compared to what I used to us. 60% of my limit has been reached already. I'm going to have to switch to an alternative now because I used to enjoy being able to cloud code tasks when I was away from my Coding Environment and on the move.

msbealo · 8 months ago

The whole affair has forced me to move to Codex CLI with local tasks using gpt-codex-mini. This is quite productive, and it was a good push. I've set up a cycling script to alternate between 2 main coder - corrector prompts as I transition my project to a different structure.
If it's useful to you, the cycling script is here. It's not perfect, but it's a lot more productive for me than Codex Cloud, which involves creating pull requests, merging commits, cutting and pasting prompts, etc., and the credit usage is pretty low.

KAMKEEL · 8 months ago

<img width="1447" height="925" alt="Image" src="https://github.com/user-attachments/assets/64ec55b2-9fd9-4126-9b1e-83f7162d40eb" />

Yeah, I have to switch to CLI, I guess. This is crazy. The weekly limit has been consumed basically already.

Dylmoco · 8 months ago

@msbealo Ive moved to cursor working locally with agents/ai, its actually really good and the pro pan is only £16 a month. You wont even notice any limits at all (they exist but hard to hit), and they reset daily, no weekly limit bs.

winnal · 8 months ago

@ae-openai

I have more evidence of some kind of bug that's massively reducing limits if they claim they did not lower them.

<img width="1257" height="832" alt="Image" src="https://github.com/user-attachments/assets/4e7268c5-3795-45c8-bc84-f431ef182bf9" />

If you look at my usage breakdown, during the period of Oct 22-25, which I highlighted, you can see that I used it for 4 days total before hitting my weekly limits. This was non-stop extensive usage of Codex exec. Then between then and now, you can see the middle period where there was a bunch of usage consecutively for a long period. This was during the week when they did rollouts of all the new dashboards and Credit system, as well as the reports of the overcharging of Codex Cloud which they reset the limits I think twice total, which was why I was able to use it for so many days without hitting any limits.

And finally now, you can see, after the limits reset again, I was able to only use Codex non-stop because I have an automated workflow orchestration pipeline, after only around 12 hours of usage my entire weekly limit was exhausted, running the exact same setup and workflow as I was running previously. this is 100% proof that the limits have been reduced by nearly 75%+ since Oct 23rd, before all the dashboard and Credit system was introduced.

To make it even more obvious, you can see there are 4 bars with high usage during Oct 22-25, which was highlighted, and only a single bar today on Nov 12-13. My entire weekly usage was used up in around 12 hours or less. Whereas previously I was able to run for the entire day for 4 days, at least 50-70 hours of usage.

Vheissu · 8 months ago

I am a Codex CLI user on the Pro plan. I have been very happy up until the past week or so where there has been a severely noticeable change in usage. I've used 10% of my weekly usage in a day despite the fact my usage of Codex CLI has not increased. If anything, the tasks I'm working on are less complicated and require less context than on previous projects/tasks I was using it on.

It's a night and day difference in that usage seems to have been reduced and no communication. So either this is a bug or it's a silent change that has not been communicated.

winnal · 8 months ago

@Vheissu 100% must be. Same thing with me. Very disappointed. I run a stable workflow that is very consistent and very repetitive doing the same tasks over and over. I used to be able to run it for nearly 24 hours per day for up to 3 full days before exhausting my weekly usage, for which I then use a second Pro account to run essentially non-stop 7 days a week.

Since just last week, after they rolled out the Credits system, my usage has dropped by nearly 90%. I can only run for a single day of 12 hours or so before my entire weekly usage is exhausted running my exact same workflow as I did previously with the Pro plan.

prabhatpushp1997 · 8 months ago

I switched to claude code. they also gave me $250 credits on claude code web. It has much more credits and it doesn't limit me a lot. I subscribed to chatgpt plus to use codex on 7th nov, but only used it once and it wiped my 5 hours credit in a single codex prompt. It was very frustrating. So later I switched to claude code. I was very disappointed as some of my friends recommended codex but my experience was very bad. they also said that it was the same case with them as well. They are also planning to switch to claude code. Literally codex is unusable at this moment.

luckysystems-nico · 8 months ago

Same problem, since GPT Codex 5.1 dropped I burst out 3-5x more of my limits with my Pro Account in Codex CLI. Before I was totally fine with the Codex Mini to handle my limits, now I will hit 100% of my pro limit within 3-4 days maximum. I think this is not intended to give us Codex Mini which is around 4 times cheaper just to cut of pro limit by multiplier 4. If it keeps like that I will consider changing my workflow back to another supplier again.

CHH-FK · 8 months ago

I also have the same issues with the Codex on Web being extraordinarily stingy with the usage. I find it baffling that OpenAI is not being more transparent nor open about this.

hadesofspades · 8 months ago

Mirroring everyone else's issues. Usage limits have been reduced to a laughable amount. Feels like a free demo rather than a paid for account.

techjoec · 8 months ago
I switched to claude code. they also gave me $250 credits on claude code web. It has much more credits and it doesn't limit me a lot. I subscribed to chatgpt plus to use codex on 7th nov, but only used it once and it wiped my 5 hours credit in a single codex prompt. It was very frustrating. So later I switched to claude code. I was very disappointed as some of my friends recommended codex but my experience was very bad. they also said that it was the same case with them as well. They are also planning to switch to claude code. Literally codex is unusable at this moment.

I moved my work to claude and increased the work immensely. Definitely getting more on $20 there than $200 here. It really is broken. Or we're ALL wrong.

aaronedev · 8 months ago

so this is not resolved or still an issue for everyone? Actually I just wanted to sub to the 200 version but now i am hesitant

winnal · 8 months ago

@aaronedev definitely not worth it since this happened. you barely get 12 hours of usage per weekly limit. it used to be 60-70 hours. nearly 80% decrease. insane. they completely refuse to address it and pretend nothing is wrong. all conveniently right after they introduced the Credits System, so that we have to pay for more usage after they rugged the usage limit.

aaronedev · 8 months ago
@aaronedev definitely not worth it since this happened. you barely get 12 hours of usage per weekly limit. it used to be 60-70 hours. nearly 80% decrease. insane. they completely refuse to address it and pretend nothing is wrong.

thanks for getting back id have to look for other alternative then i guess

KAMKEEL · 8 months ago
> @aaronedev definitely not worth it since this happened. you barely get 12 hours of usage per weekly limit. it used to be 60-70 hours. nearly 80% decrease. insane. they completely refuse to address it and pretend nothing is wrong. thanks for getting back id have to look for other alternative then i guess

Please report if you find one that's good!

techjoec · 8 months ago
> Support > Sat, Nov 8, 4:31 PM (11 hours ago) > to me > Hello, > Thank you for reaching back to OpenAI Support. > We understand you’d like confirmation about the $200 refund and additional details regarding the Codex credit. We know how important it is to have accurate information about your billing and account activity, and we’re here to assist. > After reviewing your account again, we can confirm that a $200 prorated refund was successfully issued to your original payment method, and your ChatGPT Pro subscription was canceled the same day. Please note that refunds may take up to 10 business days to appear on your statement, depending on your bank or payment provider. > At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records. We appreciate you sharing the GitHub link for context. While we’re unable to confirm any program granting automatic Codex credits, your refund has been fully processed, and no further billing adjustments appear pending. > If you have any other questions or need assistance, please feel free to reach out. We’re here to help. WTF? 1. my refund is only ~$25. 2. Apparently support STILL, as of November 8th, doesn't know a thing about the Codex credits? "we’re unable to confirm any program granting automatic Codex credits," 3. "At this time, there is no separate $200 Codex web credit or related compensation reflected in your account records." DUH that's why I opened a ticket. Still won't give me a ticket number either. @ae-openai @etraut-openai @tibo-openai @bolinfest @gpeal Anyone???
Support <support@openai.com> Tue, Nov 11, 2025 at 5:56 PM To: xxx@gmail.com Hello, Thank you for reaching back to OpenAI Support. We understand you’re requesting confirmation of the refund amount and additional details regarding the Codex credit. We’ve reviewed your account again to ensure accuracy. We can confirm that a $200 prorated refund was successfully issued to your original payment method on November 5, 2025, and your ChatGPT Pro subscription was canceled the same day. At this time, there is no record of an additional Codex credit or related compensation reflected on your account. Your refund has been processed in full, and no further billing adjustments are pending. Joe xxx <xxx@gmail.com> 3:43 PM (5 hours ago) to Support What is this case number? I need escalation to another level.

Sharing the (non) progress on my pressure on support. Sucks to see @tibo-openai and others being silenced on this; I loved when there was real interaction and seemingly pure investment and interest in fostering the platform. No hate to the folks middle-ladder doing their best, but hopefully they are also passing these kinds of long threads and pains up chain.....I suppose as a non-user I should see my way out... so probably won't say much more beyond sharing the useless responses to my "case" (which has no ticket # or anything trackable). I'm going to pester and share my case until I near my state's limit for filing the chargeback, not that I need the money, but to vocalize the rug-pull and just be a pain in the ass I guess. But I urge all that also got scammed on the credit, got scammed on the mid-subscription cycle rug-pull, and are getting flat out ignored and/or told we're imagining things, to document everything and if you decide to leave also pursue charge-backs or litigation outlets. Since support is a joke (see email chains) and the team here is under orders to, or chooses to, ignore it I think that's all we can do to get light to the situation.

I am not looking for unlimited compute resources or anything silly, and I can 110% understand reasonable limits and have decades at very large scale enterprises...But this is a real big FSCKing sheet-show the way it has gone down, and sadly I forsee bad ripples from it. But I guess the Sora gen-AI crowd really is SammyPoo's target audience, @ae-openai @etraut-openai @tibo-openai and others really have to follow their bosses' directives, I get it...OpenAI released that less than 5% of their users are coding, so I guess we are not the target market. Heard loud and clear..

/rant

prabhatpushp1997 · 8 months ago
so this is not resolved or still an issue for everyone? Actually I just wanted to sub to the 200 version but now i am hesitant

don't do it. their support is shit.😑. just go with claude code

pfahlr · 8 months ago
> so this is not resolved or still an issue for everyone? Actually I just wanted to sub to the 200 version but now i am hesitant don't do it. their support is shit.😑. just go with claude code

I did go ahead and try claude code, and in quite a few ways it works better, especially enjoying the plugins and ease of adding MCP servers... but the one way it is not better... unfortunately... is this one... wow I burned through my allotment quickly. I just noticed I had 250$ in claude code for web (they've apparently just added a web/cloud environment coding version similar to codex for web) and it expires at midnight nov 18... I think I'll be able to use it all up as I've managed to burn through $60 in credits over the past couple hours.

It could be the AI companies have decided to work together to price fix... and since all these companies have been able to replace 5-10 engineers with a single engineer running one of these tools, what's 200$ a day when it replaces 5-10 engineers that you had to pay 300-500$/day? It's still a bargain.

Unfortunately, for many a developer... it's not a bargain for us.

winnal · 8 months ago

From the latest update, the usage limits issue seems to be fixed. Usage is going down way slower now. Even when using the new max model on xhigh effort.

neo-picasso-2112 · 8 months ago

super disappointing. I am actually considering leaving this platform altogether and going open-sourced. What a disappointing rug pull from the OpenAI and Codex team.

rymatoi · 8 months ago

@etraut-openai any updates? I don't believe you marked it as a bug 3 weeks ago and it still persists 🤡

rymatoi · 8 months ago

@etraut-openai if you can't fix it you can hire me, I'll do it for you

techjoec · 8 months ago

Asked support very nicely and clearly three times now. Fully ignored for going on 7 days and @etraut-openai and others avoiding this thread completely. Jumping on other minor issues though and linking other issues to this black hole consistently however. Ah well bank charge back and burned some big customers for good. 👋🏼

etraut-openai contributor · 8 months ago

We posted an update on this topic in reddit and X a couple of days ago. Sharing here as well.

Update on Codex usage

Over the past weeks, we’ve been working to increase usage limits and fix bugs. Here’s a summary of progress:

Usage increases since Nov 1

  • Plus and Business users can send >2x more messages on average in the CLI and IDE Extension, and >3x more on Cloud.
  • Pro users can send >1.4x more messages on average in the CLI and IDE Extension, and >2x more on Cloud.
  • Enterprise and Edu plans with flexible pricing continue to offer uncapped usage.
  • How we achieved this:
  • 30% more expected efficiency (and higher intelligence too) with GPT-5-Codex-Max, compared to GPT-5-Codex and GPT-5.1-Codex.
  • 50% rate limits boost for Plus, Business, and Edu. (Priority processing for Pro and Enterprise.)
  • 30% reduction in usage consumption for Cloud tasks specifically.
  • Running multiple versions of a task (aka Best of N) on Codex Cloud is heavily discounted so that it doesn’t blow through your limits.
  • Some other smaller efficiency improvements to the prompt and harness.

Fixes & improvements

  • You can now buy credits if your ChatGPT subscription is managed via iOS or Google Play.
  • All usage dashboards now show “limits remaining.” Before this change, we saw a decent amount of confusion with the web usage dashboard showing “limits remaining,” whereas the CLI showed “limits used.”
  • Landed optimizations that help you get the same usage throughout the day, irrespective of overall Codex load or how traffic is routed. Before, you could get unlucky and hit a few cache misses in a row, leading to much less usage.
  • Fixed an issue where the CLI showed stale usage information. (You previously had to send a message to get updated usage info.)
  • [In alpha] The CLI shows information about your credit balance in addition to usage limits.
  • [Coming soon] Fixing an issue where, after upgrading your ChatGPT plan, the CLI and IDE Extension showed your old plan.

Measuring the improvements

That’s a lot of improvements and fixes! Time to measure the lifts—unfortunately we can’t just look at the daily usage data powering the in-product usage graphs. Due to the multiple rate limit resets as well as changes to the usage limits system to enable credits and increased Plus limits, that daily usage data in the past is not directly comparable.

So instead we verified how much usage people are getting by looking at production data from this past Monday & Tuesday:

  • Plus users fit 50-600 local messages and 21-86 cloud messages in a 5-hour window.
  • Pro users fit 400-4500 local messages and 141-583 cloud messages in a 5-hour window.
  • These numbers reflect the p25 and p75 of data we saw on Nov 17th & 18th. The data has a long tail so the mean is closer to the lower end of the ranges.

Bear in mind that these numbers do not reflect the expected 30% efficiency gain from GPT-5.1-Codex-Max, which launched yesterday (Nov 19th). We expect these numbers to improve significantly more!

Summary

Codex usage should now be more stable and higher than it was a month ago. Thanks to everyone who helped point out issues—we’ve been investigating them as they come and will continue to do so.

techjoec · 8 months ago

But tough luck for those of us that didn't get the credits and no comment on support refusing to acknowledge it?

etraut-openai contributor · 8 months ago

Hi @techjoec. Sorry, but I don't have any information about your particular case or your interactions with support. My understanding is that credits were given to users _as thanks for their patience_ as we investigated the issue. Reading through your posts above, it sounds like you had already canceled your subscription by that point.

aaronedev · 7 months ago

i think my limits are back to normal level, can anyone confirm the same? I just tested and ran prompts + multiple agents etc. Seems it stabilized itself

techjoec · 7 months ago
Hi @techjoec. Sorry, but I don't have any information about your particular case or your interactions with support. My understanding is that credits were given to users _as thanks for their patience_ as we investigated the issue. Reading through your posts above, it sounds like you had already canceled your subscription by that point.

Oh I did, after begging you all for review. I posted extremely detailed logs, examples, and even burned my credits _generating_ them. Full out ignored until after the 'offer' expired. Did you see how I canceled?? I am sure you are busy (no /sarcasm at all) so maybe not but I tried repeatedly to reach out to support via the dang support tool, and it refused to give me a way to make a ticket. So I literally said the below and it just did it....I was actually expecting it to give up on fluff and actually create a support ticket, and I would have stayed Plus at least. This was on Nov 4th, merely hours before you announced on reddit the credit.

The chat interaction after begging for a support ticket with the support "agent" that canceled my sub Nov 5, 2025, 5:06 AM UTC:

You said: Ok cancel my Pro The assistant said: You have been granted a prorated refund based on your usage. Your active subscription has been cancelled. The refund will show up in your bank statement within 10 days.

The specific wording you posted about the credits does not disqualify me. On the Pro plan I used Codex web during the period you stated:

To thank you for your patience, we’ve granted $200 in free credits to Plus and Pro users who used cloud tasks in the past month, valid until Nov 20.

I was patient. I posted many times to this issue AND others including verbose stats and log outputs. I begged for clarity and communication as it was impacting us. I got _nothing_ in return. Here we are over two weeks later, and I am being told that the credits were as a thank you for being patient and thus I have no claim to them? I feel I was very patient and eager to help resolve the issue for everyone, tried to provide you and the codex team with data to consider, and was just flat ignored.

To this date support has ignored my last three emails requesting escalation or even a case number and explicitly denied your claims of any credit offer even exists.

Support:

However, there is no official documentation confirming an automatic $200 Codex credit or compensation program associated with this issue.

:(

etraut-openai contributor · 7 months ago

I'm going to close this issue. There are still a few issues discussed above that we are tracking, but they are covered by other issues:
#2330
#6674
#5999