@OpenAI@OpenAIDevs@thsottiaux UPDATE: Awesome support by the @OpenAIDevs team! Thanks @reach_vb for picking this up late night PT and restoring accounts
one pending issues seems to be that restored accounts are downgraded to free tier
@IamAdiG@mont_py@OpenAI@OpenAIDevs@thsottiaux I got it, but not the subscription, at least I see my billing history, so I would expect they will fix it at some point, in the meanwhile, I can't use it
@thsottiaux Tibo - can i please get your attention to this post. Lots of people reporting that they got their accounts incorrectly deactivated due to TOS violation
https://t.co/LdRZ8BD7Xr
Woke up to both my ChatGPT and API accounts deactivated, no reason given beyond a generic "usage policy violation."
Long-time paying subscriber, using Codex to build my own software on my own machine. No VPN, no reselling, no abuse.
Appeal filed. @OpenAIDevs@OpenAI@romainhuet any chance of a human review?
.@OpenAI@OpenAIDevs my account was just banned with zero violations on my end.
My usage: normal ChatGPT in browser + Codex/OpenAI Auth inside Hermes Agent / OpenClaw.
I literally renewed my sub just a few days ago. This is clearly a bug .
Seeing others impacted too. Please help
My OpenAI Account Got Deactivated After Heavy Codex Usage
My OpenAI @OpenAI@thsottiaux account just got deactivated, and the appeal was rejected.
I mainly used Codex in VS Code for my own websites. Recently I used /goal to run a workflow that publishes articles and generates images/assets for the site.
My guess is that the long-running session may have triggered some kind of abuse or security flag. The task had been running for roughly 1 day and 2 hours, with large repo context, repeated retries, compacting errors, and lots of agent requests.
Out of panic, I continued the work from another Plus account I own, so I forgot to take a screenshot of the original task timer.
The frustrating part is I had just renewed the PRO account a few days ago, on May 30.
Has anyone else had ChatGPT or Codex access deactivated after heavy VS Code agent usage?
To my knowledge, after checking the "Why Was My Account Deactivated?" page, I did not use the account for nudity or sexual content, child exploitation, violence, self-harm, scams, deceptive behavior, hate, harassment, spam, illegal activity, or intellectual property abuse. I also did not intentionally circumvent security restrictions, share my account (except running laptop and pc on my home wifi), or share API keys inappropriately.
FYI, the reason I switched accounts was panic, not trying to bypass anything.
I thought I might lose the context in the middle of a ~28-hour Codex task, so I prioritized finishing/exporting the work first. Not ideal, but at that moment I was mostly worried about losing a full day of work with no clear reason why the account got deactivated.
This just sucks!
@Payoneer_Help@Payoneer_Latam@Payoneer since yesterday, nothing related to SMS sending or calls for security confirmations has been working, effectively blocking any operation I try to perform on my account.
Hey @OpenAIDevs, my tokens are burning way faster on Codex (Windows) than last week. Not using fast mode, on 5.4 medium. Something off or did I unlock โburn tokens aggressivelyโ? Send help before my budget vanishes.
GPT-5.4 is launching, available now in the API and Codex and rolling out over the course of the day in ChatGPT.
It's much better at knowledge work and web search, and it has native computer use capabilities.
You can steer it mid-response, and it supports 1m tokens of context.
I know this is pretty well established at this point, but Codex 5.3 is a much more effective model than Opus 4.6. I went back and forth on both for a bit, but havenโt touched Opus at all now for a full week. First model to get me off of Opusโฆ ever. Good job Codex team.
@claudeai Code is now officially free to use.
@ollama has supported @AnthropicAI's Messages API, you can now run it 100% locally with open-source models.
Get it running with this simple 5-step guide:
[๐ญ] ๐๐ป๐๐๐ฎ๐น๐น ๐ข๐น๐น๐ฎ๐บ๐ฎ
[๐ฎ] ๐ฃ๐๐น๐น ๐ฎ ๐๐๐ฟ๐ผ๐ป๐ด ๐ผ๐ฝ๐ฒ๐ป-๐๐ผ๐ฟ๐๐ฐ๐ฒ ๐ฐ๐ผ๐ฑ๐ถ๐ป๐ด ๐บ๐ผ๐ฑ๐ฒ๐น
ollama pull qwen2.5-coder
[๐ฏ] ๐๐ป๐๐๐ฎ๐น๐น ๐๐น๐ฎ๐๐ฑ๐ฒ ๐๐ผ๐ฑ๐ฒ
macOS, Linux, WSL:ย curl -fsSL https://t.co/uS3tVVUol1 | bash
Windows PowerShell: irm https://t.co/V58YRDaJdM | iex
[๐ฐ] ๐๐ผ๐ป๐ป๐ฒ๐ฐ๐
Point Claude Code to your local server instead of Anthropic's cloud:
export ANTHROPIC_AUTH_TOKEN=ollama
export ANTHROPIC_BASE_URL=http://localhost:11434
[๐ฑ] ๐ฅ๐๐ป ๐๐
claude --model qwen2.5-coder
๐ก Pro tip: Add RAG with Milvus to let your local AI search your entire codebase.
The barrier between proprietary tools and open-source models just disappeared. ๐๏ธ
Try it now!
โโโ
๐ Follow @milvusio for everything related to unstructured data!
Trailhead breaks down TUnit for .NET developersโwhat it is, why it matters, and practical steps to adopt it for faster, more reliable unit testing in your CI/CD workflow. #dotnet#unittesting
https://t.co/JnwkLuL4gz
๐ @mauiday is back! ๐ฌ๐ง
๐ Feb 6, 2026
๐ NewDay office, London
A full day of .NET MAUI, cross-platform app dev, real-world insights, and great community vibes!
You don't want to miss this!
๐ https://t.co/csnUxJaCB4