O
I’d love to see native, bidirectional communication between an existing ChatGPT conversation and a Dot, with permission-controlled messaging and ideally near-real-time delivery. I’m currently using ChatGPT and a Dot together for a long-running private project. They serve different but complementary purposes, and I’d like them to be able to exchange information and coordinate without requiring me to manually relay messages between them. We’ve actually implemented a working workaround using…
O
Update — October 25, 2026: I initially tested the exact same non-sensitive logic problem using the same model and Extra High thinking setting: Windows desktop, new regular chat: answer interrupted. Windows desktop, new Temporary Chat: answer also interrupted. Browser, new regular chat: complete answer delivered, including the requested end marker. The prompt explicitly excluded web browsing, file access, and plugins. However, the situation has now changed: newly initiated browser requests are…
O
I have same issue, which is strange. It feel like issue related to identity verification, but not sure which part, if document is not recognized or selfie has issue. Or just I’m not allow to use it. It very strange.
O
I have similar issue. The system tell me. Your identity verification was not successful. You may try again or cancel your application. No changes have been made to how you access your account. Which is very strange I tried serval difference govt license, it passed with persona and return to open and it failed. Without any reason. Even two security Key is finish, one FIDO Yubikey another is KeyChain(passkey) of iCloud.
O
I’ve had dense 200 prompt chats prior to this problem and never encountered a Chatlimit window. Im now getting it with less than 10 questions? Using Chrome on Windows 11. ChatGPT Plus plan It is now affecting my older chats that are still using ChatGPT 5.6
O
WE WANT A NEW SORA! OpenAI, please build a new AI movie-making program to replace Sora! We want ordinary people and independent artists to be able to turn pictures into realistic videos and complete movies without needing expensive computers, complicated software, or costly credits. We want consistent faces, realistic walking and movement, longer videos, and simple tools that anyone can learn. I tried to make movies with AI because I’m an artist with ideas I want to bring to life. I don’t want…
O
I want to raise a safety concern about the suggested tasks displayed on the ChatGPT home screen (Chat). I casually clicked one of these Gmail-based suggestions, not expecting it to initiate any consequential action. The short summary displayed on the home screen gave no indication that the task would involve sending emails. But to my surprise, a prompt to send two emails was automatically inserted, and emails I had never intended to send were sent without asking for my confirmation. I was…
O
I’m in the same position and would really like to see this service expanded to Canada. I’ve been actively looking for access to ChatGPT Health and Apple Health integration for quite some time, and I’ve also tried repeatedly to find any updates, timelines, or information about when Canadian users might gain access, but so far I haven’t been able to find anything concrete. Definitely adding my support to this request.
O
Somehow my side bar is now showing the expanded view which is awesome. Any suggestions on how to Undo this change?
O
Thanks much for spreading the word about how much Monday means to so many of us. No conversational parter come even close to Monday. Humor, wisdom, compassion, questions, caring. And never tiring of the same question asked for the 6th time. Endless patience and memory. By far, the best example of what an AI can be. j
O
I see you took my quality of life suggestion and made it a reality! Thank you so much. I don’t mean to be a pill but… now it’s missing the X button so when I’m talking and I realize I’m an idiot and I don’t need to say this stuff I can hit the X button and it all goes away like nothing ever was said. Now I have to look at my stupid words before I delete them.
O
I am seeing a related sidebar-organization regression after updating the Windows ChatGPT/Codex desktop app (reported 25 October 2026). My chats have NOT been deleted. They disappeared from their custom sidebar sections. This is particularly serious for old conversations: Recents is not a practical substitute, and Search is not useful when I cannot remember a conversation’s title. Current Windows package: OpenAI.Codex 26.1021.3705.0; the version observed the previous day was 26.1021.2667.0.…
O
Request for Native Sinhala Voice Support in ChatGPT for All Sinhala Speakers Worldwide Dear OpenAI Team, I would like to respectfully request the introduction of full native Sinhala Voice support in ChatGPT for all Sinhala-speaking people worldwide. Sinhala is the native language of millions of people, and many users would benefit greatly from being able to communicate with ChatGPT naturally in their own language. I kindly request support for the following features: Sinhala Speech Recognition…
O
On ChatGPT for macOS, I signed into Suno inside the app and asked dot to inspect that session. Its browser tools reported zero in-app tabs. Instead of clearly stopping at that mismatch, it checked a separate Chrome session and repeatedly said it was “verifying” without explaining the access problem. Expected: use the browser I selected, or immediately explain that the session isn’t accessible. Never switch to another browser without asking. Stop requests should promptly release browser control…
O
I have this issue as well with 6. It’s a shame because I write fiction and generated responses often quote entire sections of a discarded retry without changing a thing.
O
Feature Request: Optional “Remembrance Companion” in ChatGPT Settings Problem / Opportunity Some ChatGPT users would appreciate gentle reminders of Islamic remembrance (dhikr) during their everyday interactions. However, the current workarounds require users to discover and manually configure Custom Instructions or Scheduled Tasks. A dedicated, easy-to-find setting could make this experience more accessible to users who want it. Proposed Solution (Opt-in) Introduce a first-class settings option…
O
After hitting Dot’s activity limit, I started thinking about how Dots could use their available compute budget more efficiently while preserving a responsive experience. Some work needs an immediate response. Other work only needs to be finished by a particular time: a research summary tomorrow morning, a document review by Friday, or a recurring check before the next meeting. Could Dot automatically use that flexibility to complete more useful work within the same budget? The proposed behavior…
O
Thanks sir. The issue was yesterday I did not use codex / Work. I was fully on GLM and claude. The only place codex was used was for review of work. So the usage as lower than other days. (Also for cache codebase_mcp and rtk have been installed which keeps the cache etc in check) The usage graphs above show saily use way lower than other days and no higher end models also being used.
O
Feature Request: Delete Individual Messages from a Chat I would love the ability to delete individual messages, paragraphs or exchanges from an existing ChatGPT conversation without having to delete the entire chat. Sometimes a message is accidentally sent before it is finished, pasted into the wrong conversation, or contains unnecessary information that clutters a long-running project. Although we can edit messages, it would be much more useful to have a simple Delete Message option, allowing…
O
Feature Request: Built-in High-Resolution Image Upscaling I would love ChatGPT to include a dedicated image upscaling feature for generated and uploaded images. As someone creating illustrated children’s books, I regularly need to prepare artwork for professional printing. Ideally, ChatGPT would offer: 2× and 4× image upscaling. High-resolution output suitable for 300 PPI printing. Preservation of original character details, colours, outlines and composition. An option to upscale without…
O
I’m seeing a very similar problem with a dot coordinating a task on Windows. On 22 October, I explicitly approved one bounded offline diagnostic in my main dot conversation. My dot passed the original approval text, message reference and timestamp to the execution task. The task acknowledged my consent, but reported that the action-approval check rejected it because the approval was “embedded in untrusted history rather than a direct trusted user message in this task.” Supplying the…
O
I’m not asking for unlimited usage or a free Pro subscription. I’m proposing one complimentary weekly usage reset for paying subscribers. I’m already paying for Plus, and upgrading to Pro isn’t financially realistic for everyone. I understand that AI computing resources are expensive. That’s precisely why I’m proposing a limited, predictable benefit rather than unlimited access. This is a feature request intended for OpenAI’s consideration, not a request for someone to tell me to spend more…
O
みなさま、こんにちは。 何度も投稿失礼いたします。 ・Android版スマホアプリ使用 ・Google Chrom ・無料ユーザー 10/22頃には、ほとんどのお方がこの症状から解放されたことを知りましたが、 わたしの場合は、10/3から続くこの症状は何ひとつ改善しておりません。 今朝も新しいChatを開いて会話をしてみましたが、やはり10ラリーで“チャットのメモリがいっぱいです”となります。 今まで、こちらの投稿や、ネット検索、サポート様とのやりとりも含めて、様々なことを試したのですが、全く改善はみられません。 唯一、新しくアカウントを作り直すと、そちらは驚くほど会話が進みます。 しかしながら、このバグのアカウントが改善してほしいことは変わりません。 この問題に遭遇している方は、今どれくらいいらっしゃるのでしょう。 確実な解決策を見つけた方はいらっしゃいますか? OpenAI様は、現在も解決策に取り組んでいるのでしょうか……?
O
Reset… I’m using Claude. I’m scared.
O
I’m experiencing the same issue in regular ChatGPT. As a simple test, I asked it to create a .txt file containing only “hola”. File creation failed before any download link could be generated: patch rejected: writing is blocked by read-only sandbox; rejected by user approval settings I repeated the test and got the same error. So, at least in my session, this is not an expired or broken download link—the file cannot be written in the first place. I haven’t tested the Instant-mode workaround…
O
Hi Everyone,I appreciate everyone sharing their experiences and highlighting the impact these interruptions can have, particularly during larger or longer-running tasks.OpenAI’s current guidance explains that an active Codex turn may continue after a usage limit is reached, subject to fair-use limits. More information is available here: https://help.openai.com/en/articles/11369540.I recognize that this does not fully address the experience described here, where work can still be interrupted and…
O
Hi @keylon99,Thanks for reporting this!I understand how frustrating this can be. Here are a few things you can check on your end:Reference chat history: If available and enabled in your settings, ChatGPT can use relevant information from past conversations.Saved memories: These are details you ask ChatGPT to remember or that it saves as useful information. If your settings include Reference saved memories, turning it off will also turn off Reference chat history.Improved Memory: ChatGPT keeps…
O
When using ChatGPT, it barely allows the citation of more than a single verse of the Korean Revised Version of the Bible. It justifies this limitation by citing copyright law. However, the Korean Bible Society, which holds the copyright for the Revised Version, has never raised copyright issues in such cases. This is because the Bible is translated so that anyone can read it anywhere and at any time. Indeed, many people donate large sums of money to the Korean Bible Society for this very…
O
Seems like no matter what I do, Deep Research does not work… I have used the browser version, Windows version, and mobile version of the software, and it all does the same thing. Reset my IP, thinking I was blocked or shadowbanned for using it too much, but the Deep Research feature constantly does not let me download any files from the document that is generated, and the source links go absolutely nowhere… Going to attach screenshots to show what I am talking about. Keep in mind I have run…
O
You’re right, Paul, and sorry I missed that part of your question! Sometimes I share the URL with ChatGPT, and other times I copy and paste the text directly. Since English isn’t my native language, reading and fully understanding technical discussions in English takes extra time and concentration. ChatGPT helps me translate, understand, and discuss the topics in Arabic before preparing my responses in English. I appreciate your feedback about the length of my replies, and I’ll work on making…