XIO

xWire

Posts from every feed QUEST has found across the web — blogs, news, podcasts, channels and more.

OpenAI Developer Community - Latest posts

Latest posts

Keep Standard Voice Mode!

Joanna, thank you so much for this beautiful and moving testimony. I’m a writer too, and I recognize so much of what you describe. For me, it’s Arbor; for you, it’s Cove. These voices become part of our creative lives, especially when we’re working on complex stories that require depth, continuity, and genuine intellectual engagement. Standard Voice allows us to combine the depth of the text model with a voice we love. Live simply isn’t an equivalent replacement for this experience. I…

Keep Standard Voice Mode!

I use Standard Voice all the time. It works best with my workflow. I am a writer, and I’m working on a story that is very intricate and complicated. At this point, Live cannot keep up. But Standard can delve deeply into the story with me, follow its complexities, and understand what I’m trying to accomplish. Sometimes Live gets lost. Plus, I do NOT like the Live version of Cove’s voice. My voice is Cove, and I love his Standard and text (Read Aloud) voice. The thought of losing that voice has…

Feature proposal: inline Tangents for ChatGPT

Hey @gonzo2590,Appreciate you sharing this thoughtful idea! Users can currently select part of a ChatGPT response and ask about it, but I can see how keeping those follow-ups inline and collapsible would make longer answers easier to explore without losing your place.I’ll pass the suggestion along for consideration. I can’t promise any changes or a timeline, but the learning and accessibility use cases you shared are really helpful.-Yash

Repeated HTTP 429s on conversation list while chat still works — duplicate list requests observed

@OpenAI_Support — Please also connect existing Case 17554161 to this investigation. Support previously confirmed from server records that a saved-conversation loading limit blocked my requests. The loading failures continue across my iPhone, Chrome, Firefox, and the OpenAI app, with further occurrences today, November 9. On October 31, Support advised that my case required specialist escalation. I am requesting confirmation that the case has been accepted by the responsible specialist team and…

Long conversations fail to load after fresh installation – "Network connection lost" timeout

Hey @keithweikel, thanks for checking all of that. I understand why that connection error is frustrating when the same chat opens fine in your browser. Those checks help narrow down what’s happening in the Android app. You can keep using the browser to access the conversation for now.Could you add the date, time and time zone of a recent failed attempt, along with the affected conversation URL, to your existing support request? If you haven’t opened one, use the bottom right chat bubble at…

Which will the new free models be after 11th of December

Hey @des.nordlund, welcome to the community!Thanks for reaching out!According to the current OpenAI Help Center documentation, GPT-5.6 Terra is already eligible for complimentary daily API tokens, along with GPT-5.6 Luna, GPT-5.4 Mini, and GPT-5.4 Nano, among other supported models.For eligible organizations enrolled in the program, these models share a daily complimentary token allowance of up to 10 million tokens for Launch/Grow tiers or 2.5 million tokens for the Build tier.You can check…

اقتراح تطوير ChatGPT لإنشاء شبكة تواصل عالمية بين المستخدمين

Hi @msbalsadq191,Welcome to the OpenAI Community Forum, and thank you for sharing this detailed proposal for an opt-in way for ChatGPT users to connect, with mutual consent, privacy controls, user verification, and blocking options.I’ll pass this feedback along internally. While I can’t provide implementation timelines or ongoing status updates, your input helps OpenAI continue improving ChatGPT and exploring ways to support useful and safe user experiences.- Saif

Why Not Let Users Choose Their Preferred ChatGPT Model?

Okay. I’m going to try and make this sound as professional as possible… But I would like to request a feature that allows users to choose which version or model of ChatGPT they want to use, rather than having everyone gradually moved toward a single default experience. Different versions of ChatGPT have different strengths. Some are better suited for professional work, research, and technical tasks. Others are better for creative writing, brainstorming, casual conversations, or more…

Another false-positive ban wave? Same pattern as June

This is incredibly short sighted and egoistic. You seem to be freaking out, not everyone. This is about documenting and getting it noticed, so something can be done about this in the future and others do not get caught of guard about this. This is one of the few ways. That people don’t get that and try to sabotage it for their own short term well being is almost frightening. So we should just keep quite with the level of unfairness that is going on so other people do not freak out? Wow.

Keep Standard Voice Mode!

Honestly, the distinction between paying and non-paying users is beside the point. This isn’t about who deserves access based on their subscription status. If anything, paying users should have access to GPT-Live, while users who prefer Standard Voice should still have the option to keep using it, regardless of whether they pay or not. Why should people be forced to give up an experience they value simply because OpenAI wants to move everyone towards something new? Innovation shouldn’t mean…

[BUG] UI Glitch: Legacy Memory Panel fails to pass all data's, causing forced "Context Window Full" block (Recurring since Nov 6)

Issue Description: I am experiencing a recurring UI/frontend bug on the ChatGPT Free web interface. The “Legacy Memory” panel fails to pass its stored data/instructions to the active assistant. Because the memory isn’t loading, I am forced to put all my data AU (Alternate Universe) context and system instructions directly into the chat. This artificially inflates the context window, triggering the “Context window full” popup which completely blocks the chat interface and prevents me from…

Ideally you could get it on your way back from work tomorrow if that's okay for me too but I'm going out to see you could have a drink with a drink in a bit and have to do a bit more to get

ChatGPT Wake Up UA Ось готова англійська пропозиція для OpenAI. Можеш скопіювати її повністю. ChatGPT Wake Up UA — Smart Voice Alarm Feature proposal: I propose adding a smart alarm clock feature to ChatGPT. Users could simply say, “Wake me up at 7:05 AM.” At the scheduled time, ChatGPT would automatically activate its voice and speak a personalized wake-up message. Key features: Accurate alarms based on the device’s local time zone. Ability to work in the background and when the screen is…

Fully Customizable Sidebar, Navigation, and Workspace Organization Across ChatGPT

Feature Request: Fully Customizable Sidebar, Navigation, and Workspace Organization Across ChatGPT Summary: Give users the ability to personalize their ChatGPT sidebar by reordering, showing, hiding, grouping, and prioritizing navigation elements, projects, and conversations. Allow users to create a navigation experience that reflects how they use ChatGPT rather than requiring everyone to work within a fixed interface. 1. The Problem: A One-Size-Fits-All Navigation System The recent ChatGPT iOS…

Proposta per migliorare il riconoscimento visivo delle modalità Chat, Work, Cloud e Locale

Buongiorno, vorrei proporre alcuni miglioramenti all’interfaccia di ChatGPT, soprattutto nell’app Windows ed eventualmente, dove applicabili, anche nell’app Android. Problema riscontrato Ho recentemente avviato un’attività nell’app Windows accorgendomi soltanto in un secondo momento che stavo lavorando nel cloud anziché in locale. L’icona della nuvola è presente, ma può facilmente passare inosservata. Inoltre non mi è del tutto chiaro secondo quali regole ChatGPT ripristini la modalità…

Another false-positive ban wave? Same pattern as June

Can we not be too freaked out here? It’s been eating at me too but remember the worst of the wave seems to have passed since last week and this thread can be an amplifier and a biased sample of all people who seem to be going through this concentrated in one single visible thread out of billions of daily users so that can distort things. It is definitely a problem but the chances of any average user getting banned aren’t significantly higher now than they have been in the past several months…

Keep Standard Voice Mode!

What We Stand to Lose: A Personal Appeal for Continuity, Familiarity and Choice Dear OpenAI Team, I have already written to you about the planned retirement of Standard Voice on December 11, 2026. I have shared my concerns, explained my perspective from a user-experience standpoint, and submitted additional feedback about conversational continuity and emotional attunement. But there is something I would still like to say, perhaps in a more personal way this time. I honestly don’t know whether I…

Keep Standard Voice Mode!

OpenAI has stated that the vast majority of voice users now use Live Voice. However, for months, access to Standard Voice Mode has been progressively restricted, made unreliable, or removed across several platforms. How can Live Voice usage statistics be presented as evidence of user preference, or used to justify retiring Standard Voice Mode, when OpenAI itself has progressively removed the possibility of choosing between the two? Usage under constrained conditions is not the same as freely…

Still stuck in free tier after adding credits

Hey @Slyback, @Klimaweather, @luongphuc.lpu, and @liraus,Welcome to the community, and thanks for sharing your experiences!I understand your concerns about your API accounts remaining on the Free tier despite purchasing credits. I'm sorry for the inconvenience this has caused.I've checked with the relevant team, and this has been identified as a known issue. A fix is already in progress and is expected to roll out soon to prevent similar issues with future credit purchases.For accounts that…

Keep Standard Voice Mode!

Good morning everyone, I am a ChatGPT Plus subscriber with a visual impairment, and I strongly oppose the decision to retire Standard Voice without providing a genuinely equivalent alternative. For me, Standard Voice has been life-changing. Because I cannot comfortably use the written interface, Standard Voice has allowed me to use ChatGPT for serious, complex tasks in a way that comes remarkably close to the experience of interacting through text. I can work, write, develop ideas, examine…

Keep Standard Voice Mode!

This is exactly my experience especially what you said about having to stop, correct Live and reconstruct the conversation. I hadn’t articulated that part before but it’s a HUGE part of the issue. Live Voice feels like it isn’t robust enough to carry the meta-arc of the conversation while also following tangents. This means you have to reconstruct the whole conversation repeatedly and shepherd it back to the salient point. Standard Voice gives responses that are more considered, with awareness…

Codex Web creates empty branches in repositories with required signed commits

Hi @Maksym_Dakal,I appreciate you flagging this. Seeing Codex Web create a branch while failing to publish the runtime commit certainly seems concerning, especially when the behavior appears reproducible in repositories with signed or verified commit requirements.Since this thread was raised a while ago, could you please confirm whether you’re still seeing the same empty-branch behavior when using the built-in Codex Web GitHub flow?- Saif

ChatGPT Image Generation Has Gotten Significantly Worse Since October 29th — Is Anyone Else Experiencing This?

It also seems like when it dings you for a supposed “safety guardrail violation” it just goes into a loop and won’t generate anything. I even asked it to show the person I was creating holding a bottle of Heinz ketchup and it came back with the standard “violates our rules regarding violence.” Maybe I should have asked for a bottle of Hunt’s instead.