"You've hit your usage limit." Despite 10% Rate Limit Usage Remaining

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

• #10663

Powered by Codex Action

Thread ID:

9888a61d-81e0-44b2-97a5-e20f782f112d

same issue here. have 21% remaining when opening a new session in codex CLI and the website usage page shows 21% but when i try to send any message it get the usage limits message.

Thread ID: 019c7a95-753c-7df2-9c77-3d122f357ef4

@justinTM, I can't find your thread ID in our database. Did you use /feedback to upload the logs and session details? If so, it appears it didn't upload successfully. If you're able to repro in a new session, please use /feedback to upload the logs and session details and post the thread ID here.

@TJZine, I see your thread ID in our database. I quickly looked at your logs. I'm guessing that you're using the experimental subagents feature? Have you overridden the default value of agents.max_threads in your config? You triggered a 429 (Too Many Requests) error from our service, which can happen if you set max_threads beyond the default value.

@etraut-openai I do have multi_agent feature set to true in the config but no other settings related to it like agents.max_threads. I just tried setting multi_agent to false but I still get the same usage error trying a test message again in a fresh chat.

Can you /feedback and share the contents of your config.toml files?

okay submitted through feedback, and pasted my current config with the bug report. New thread ID: 019c7ca0-4561-7f43-b66c-a24d2cdc2a10

@TJZine, the logs show that you are at 0% remaining for your weekly utilization. Where are you seeing that you have 10% remaining? Is that appearing in the app?

@etraut-openai I see 21% remaining on the usage webpage as seen here.

<img width="1222" height="1614" alt="Image" src="https://github.com/user-attachments/assets/7572a541-50c9-4128-8234-65414f254bde" />

I also see it when opening a new chat in codex and using /status

<img width="844" height="492" alt="Image" src="https://github.com/user-attachments/assets/8615c8ad-0402-4910-949d-9527e7adf2f1" />

But when I send anything and run status again it then shows 0%

<img width="813" height="398" alt="Image" src="https://github.com/user-attachments/assets/80e5ff97-6a72-4e37-8e2c-459513aed20f" />

Not sure exactly whats going on but I should have the 21% left based on my monitoring on my weekly usage and a 20% drop is pretty big/noticeable especially with the 2x limits.

@etraut-openai thanks for your reply -- just re-uploaded. ID:

ef7e20d7-5c56-48eb-b3eb-ec63fcedf674

contents of ~/.codex/config.toml:

model = "gpt-5.1-codex-mini"
model_reasoning_effort = "medium"
personality = "pragmatic"

[features]
shell_snapshot = true

[projects."/home/ec2-user"]
trust_level = "trusted"

@justinTM, I can't find your thread ID in our database. Did you use /feedback to upload the logs and session details? If so, it appears it didn't upload successfully. If you're able to repro in a new session, please use /feedback to upload the logs and session details and post the thread ID here. @TJZine, I see your thread ID in our database. I quickly looked at your logs. I'm guessing that you're using the experimental subagents feature? Have you overridden the default value of agents.max_threads in your config? You triggered a 429 (Too Many Requests) error from our service, which can happen if you set max_threads beyond the default value.

Hi @jif-oai @etraut-openai — I’m seeing the same issue as reported here.

What I see

• The web Usage Dashboard shows I still have quota remaining (5h limit and weekly limit).
• codex CLI /status also shows remaining usage (e.g., 100% left for 5h and ~42% weekly left, with the reset time).
• But as soon as I send any message, I immediately get:

“You've hit your usage limit. Upgrade to Pro … or try again at Feb 25th, 2026 8:07 PM.”

After that, /status may flip to 0% remaining.

Screenshots

• Web usage page showing remaining quota:

<img width="1578" height="590" alt="Image" src="https://github.com/user-attachments/assets/0d6b6cfa-b24a-4509-ace7-0d2adc36d451" />

• CLI /status before sending a message:

<img width="656" height="247" alt="Image" src="https://github.com/user-attachments/assets/447d7e25-82b7-47b3-89e8-56b30abafbcd" />

• Error message after sending a message (and /status after):

<img width="1188" height="118" alt="Image" src="https://github.com/user-attachments/assets/befdedc8-07e5-455a-9be2-f09d39cb6637" />

<img width="604" height="239" alt="Image" src="https://github.com/user-attachments/assets/c7a6a198-11cc-4652-855d-366d7e30806d" />

Repro steps

1. Open a fresh Codex session (or new chat in the IDE extension/CLI).
2. Run /status → shows remaining usage.
3. Send any prompt → immediately receives the “hit your usage limit” message.

Logs / Thread ID
I reproduced in a new session and ran /feedback to upload logs.

Thread ID: 019c8b0b-8984-7e93-80ad-0f49cdcc80fb

Environment

• Subscription: Plus
• Client: Codex CLI version v0.104.0

<img width="1297" height="540" alt="Image" src="https://github.com/user-attachments/assets/6b2a5f10-c30d-4887-bf96-92245c3b2bc1" />
Thread ID: 019c8e22-5627-79d1-8298-8c02db63d788

Additional debug info:

Thread ID: 019c8f92-d2c3-7ee3-9cd5-d050276b22a6

Feedback has been uploaded via /feedback.

Repro:

• /status shows included quota remaining (38% weekly)
• codex exec "Say pong" -> usage limit error
• codex exec --model gpt-5.3-codex-spark -> works
• Same account works on another machine

This appears to be quota bucket misrouting specific to this device/session.

same issue. ID: 019ddcd1-396a-77f1-94a6-47a14518ae5b

/Status result:

<img width="822" height="192" alt="Image" src="https://github.com/user-attachments/assets/89690418-1f36-4275-920e-31660b25d02d" />

Hi. Same here. Thread ID:
019dd506-40c6-78c1-a33c-5f620d51ec23
Thanks

╭─────────────────────────────────────────────────────────╮
│ >_ OpenAI Codex (v0.135.0)                              │
│                                                         │
│ model:       gpt-5.3-codex xhigh   /model to change     │
│ directory:   /home/…/launch-fix-autopolot-observation-5 │
│ permissions: YOLO mode                                  │
╰─────────────────────────────────────────────────────────╯

  Tip: New Use /fast to enable our fastest inference with increased plan usage.

/status

╭────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│  >_ OpenAI Codex (v0.135.0)                                                                            │
│                                                                                                        │
│ Visit https://chatgpt.com/codex/settings/usage for up-to-date                                          │
│ information on rate limits and credits                                                                 │
│                                                                                                        │
│  Model:                       gpt-5.3-codex (reasoning xhigh, summaries auto)                          │
│  Directory:                   /home/tools/oh-my-codex.omx-worktrees/launch-fix-autopolot-observation-5 │
│  Permissions:                 Full Access                                                              │
│  Agents.md:                   <none>                                                                   │
│  Account:                     commonvibe@protonmail.com (Pro Lite)                                     │
│  Collaboration mode:          Default                                                                  │
│  Session:                     019e7d7d-2190-74d3-b76f-7f39dfdf64eb                                     │
│                                                                                                        │
│  Context window:              100% left (0 used / 250K)                                                │
│  5h limit:                    [███████████████████░] 94% left (resets 13:26)                           │
│  Weekly limit:                [░░░░░░░░░░░░░░░░░░░░] 0% left (resets 13:36)                            │
│  GPT-5.3-Codex-Spark limit:                                                                            │
│  5h limit:                    [████████████████████] 100% left (resets 18:03)                          │
│  Weekly limit:                [█████████████░░░░░░░] 66% left (resets 16:49 on 4 Jun)                  │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────╯


› ping


⚠ Skill descriptions were shortened to fit the 2% skills context budget. Codex can still see every skill, but some descriptions are shorter. Disable unused skills or plugins to leave more room for the rest.

• SessionStart hook (completed)
  hook context: [Execution environment]- surface: attached tmux runtime - tmux- omx runtime surfaces: omx team, omx hud, and omx question are directly usable in this session- omx question: visible temporary
  renderer available from the current pane; primary success JSON is answers[]

■ You've hit your usage limit. Visit https://chatgpt.com/codex/settings/usage to purchase more credits or try again at 1:36 PM.

gpt-5.3-codex xhigh · /home/tools/oh-my-codex.omx-worktrees/launch-fix-autopolot-observation-5 · fix/autopolot-observation-5 · Context 100% left · Context 0% used · 250K window · 0 in · 0 out · 5h 94% left 

■ You've hit your usage limit. Visit https://chatgpt.com/codex/settings/usage to purchase more credits or try again at 1:36 PM.
To continue this session, run codex resume 019e7d7d-2190-74d3-b76f-7f39dfdf64eb

I'm experiencing the same issue with a ChatGPT Plus account.

Details:

• Active ChatGPT Plus subscription (auto-renewal date is visible in billing settings).
• Codex usage page is accessible and shows available quota (5h limit with ~99% remaining).
• Codex is available in my account.
• However, when attempting to use Codex, I receive the following error:

"The 'gpt-5.3-codex' model is not supported when using Codex with a ChatGPT account."

This does not appear to be a quota or subscription issue, since usage limits are still available. It seems like an account entitlement, model mapping, or backend synchronization problem.

Has anyone found a workaround, or can the team confirm whether this is a known issue affecting ChatGPT Plus users?

■ You've hit your usage limit. Upgrade to Pro (https://chatgpt.com/explore/pro), visit https://chatgpt.com/codex/settings/usage to purchase more credits or try again at 4:01 AM.

/status

╭─────────────────────────────────────────────────────────────────────────────────────────╮
│  >_ OpenAI Codex (v0.137.0)                                                             │
│                                                                                         │
│ Visit https://chatgpt.com/codex/settings/usage for up-to-date                           │
│ information on rate limits and credits                                                  │
│                                                                                         │
│  Model:                       gpt-5.5 (reasoning medium, summaries detailed)            │
│  Permissions:                 Workspace (Approve for me)                                │
│  Collaboration mode:          Default                                                   │
│  Session:                     019e9990-4c2b-77a2-b9da-9a86d1a06064                      │
│                                                                                         │
│  Context window:              30% left (185K used / 258K)                               │
│  5h limit:                    [███████████████░░░░░] 76% left (resets 04:01)            │
│  Weekly limit:                [███████████████████░] 96% left (resets 23:01 on 12 Jun)  │
│  GPT-5.3-Codex-Spark limit:                                                             │
│  5h limit:                    [████████████████████] 100% left (resets 05:24)           │
│  Weekly limit:                [████████████████████] 100% left (resets 00:24 on 13 Jun) │
│  premium limit:                                                                         │
╰─────────────────────────────────────────────────────────────────────────────────────────╯

<img width="1485" height="368" alt="Image" src="https://github.com/user-attachments/assets/b0d57c46-d14e-4b6e-8bf1-27673efd2c82" />

...and it is 1:58 a.m. so the date it wants me to wait for is in the past.

I've just purchased my sub BTW.

Session ID: 019ea375-8696-7462-ae1c-04fbe5b702eb

Same problem here, encountered today at 3% remaining with a Plus subscription. Session ID 019ecadb-94a8-7c42-87fc-e8bbc536d99d.

<img width="1085" height="252" alt="Image" src="https://github.com/user-attachments/assets/d0476165-8ad8-4780-b953-1ba5b620f09f" />

EDIT: sending a message with /feedback as the only content in the Codex app (MacOS, version 26.609.41114) actually successfully triggered the prompt that had failed multiple times earlier (also after reopening the app multiple times and double-checking my remaining usage), and made it run to completion.

Same here, I just purchase the sub.
100% left but hit the usage limit already.
session ID: 019e7c59-b56f-74d0-aa3c-a5474aca31da

<img width="3168" height="1910" alt="Image" src="https://github.com/user-attachments/assets/88d41dd4-f96b-4d1c-8574-97e12e872ed3" />

I hit a very similar issue after upgrading from Plus to Pro Lite.

Symptoms:

• Codex CLI /status showed Account: Pro Lite.
• /status showed remaining quota, e.g. 80% 5h limit and 97% weekly limit.
• Any prompt still failed with:

You've hit your usage limit. Upgrade to Pro ... or try again at ...

• The issue persisted in the same session after the plan upgrade.
• gpt-5.3-codex-spark also appeared in /status with 100% quota, but selecting it failed with:

The 'gpt-5.3-codex-spark' model is not supported when using Codex with a ChatGPT account.

Fix that worked:

• codex logout
• renamed %USERPROFILE%\.codex to .codex_BACKUP
• ran codex login again
• started a fresh Codex session with gpt-5.5

After recreating the local .codex directory, gpt-5.5 prompts worked again.

This suggests stale local Codex state after a Plus -> Pro Lite upgrade can cause false usage-limit failures even when /status and the usage dashboard show available quota.

Same here. Absolutely ridiculous. Used /feedback multiple times. I've paid for a service and it isn't being provided.

Feedback ID 019ed59b-8545-7473-b21e-dc463408ca46

Same, just bought pro version and I am rate-limtied now. used limit reset, now +3 days to rate limit 🫤
feedback id 019ee944-fc2a-7740-a8d5-7336a790000b