Scheduled Tasks – Model Selection
I have the same issue. I need to run a scheduled task as GPT-6.1 Sol with High reasoning, but it use 6 Instant. I would like to congfigure the reasoning level for the scheduled task
Posts from every feed QUEST has found across the web — blogs, news, podcasts, channels and more.
I have the same issue. I need to run a scheduled task as GPT-6.1 Sol with High reasoning, but it use 6 Instant. I would like to congfigure the reasoning level for the scheduled task
I would like to suggest an AI-powered platform that connects people with business ideas to potential investors. Anyone should be able to submit an idea for free, without needing money, a company, or personal connections. AI could evaluate ideas, explain their strengths and risks, and help match them with interested investors. Both idea creators and investors would use the basic platform for free. Revenue could come from advertising. To protect creators, each idea would have a unique code. Only…
Yeah… this is exactly what makes it so draining. The same model can treat the same content completely differently from one moment to the next, and when you ask why, it often can’t even make sense of what changed. Imagine having someone in your life who treats you differently every day, whose mood and boundaries constantly shift. How long would you keep them around? That’s honestly what this product behavior feels like to me.. unstable, exhausting, and pretty toxic to deal with mentally.
Thanks, let me know if you need any info from me.
Dear OpenAI Product Team, I would like to propose an idea that could help make ChatGPT more accessible to people around the world, particularly in lower-income countries. The Problem Millions of people want to use AI for education, business, research, creativity, and personal development. However, subscription costs can be a significant barrier for users in countries where average incomes are much lower. Many people cannot afford a paid ChatGPT subscription, even though AI could provide them…
Treating adults like adults clearly isn’t impossible. The infrastructure exists, age verification exists, and competitors are already doing it. At this point, it’s a product choice. What I also dislike is using live users as unpaid QA through constant A/B tests and rollouts. We’re left to find the regressions, document them, and deal with the stress and mental toll ourselves. Ha, it would be nice if all of this actually ended up benefiting users in the end.
By chance have you made a script that goes and checks space used and runs a cleanup with the persistent pointers? I use custom hybrid persistent memory on my AI’s and LLM’s.
To be honest, at this point I’m just waiting for Muse Wonder to launch. From the way OpenAI has been handling things lately, it really doesn’t look like user complaints, concerns, or long-standing problems are much of a priority. Now they’re talking about bringing Chat and Work together too, and people are already worried that normal Chat usage will eventually get dragged into the same kind of limits and quota system. And personally, I cannot stand the GPT-6 models. For writing, creativity,…
Just ran into the same thing! Cheers for flagging
Obligatory, “restart the timer reset countdown with Luna (for if I were to unexpectedly use Codex)” (I guess not zero trees, or maybe four, or maybe twice that)

That sounds really frustrating, especially with years of worldbuilding in those chats. A couple of things that helped other people while this gets fixed: “Memory full” is about the saved memories list, not the chat itself. Settings > Personalization > Manage memories lets you delete older or duplicate entries, which usually clears the warning right away. Small, repeated facts tend to pile up there. For continuity, keep a “series bible” outside the chat: a document with characters, relationships…
Dear OpenAI Product and Research Teams, I would like to propose an opt-in, anonymized knowledge base of verified engineering solutions, built from real-world technical investigations conducted with ChatGPT. The problem Engineers worldwide repeatedly investigate similar industrial equipment failures. Significant effort goes into analyzing PLC programs, historical trends, sensor measurements, equipment documentation and possible causes. Once a problem is solved, the resulting knowledge often…
Please add support for clickable hyperlinks that navigate within an individual ChatGPT response. Specifically: A Back to top link to return to the beginning of the response. A clickable table of contents that jumps directly to sections within the response. Back to contents links at the end of individual sections. Consistent functionality across the ChatGPT website, iOS and Android apps, preferably using standard Markdown anchor links. This would be particularly valuable for long technical…
Dear OpenAI Team, I am a regular user of ChatGPT, Work, and Codex, tools that I greatly appreciate and that have enabled me to undertake projects I probably would never have been able to accomplish on my own. With their help, I have developed several macOS applications, including PhotoFil, a program designed to organize photographs chronologically, particularly old scanned photographs. Throughout these projects, I have encountered some difficulties related to conversation organization and…
Proposal: One-Click AI Software Verification and Signing for Independent Creators Making software creation and distribution accessible to everyone AI is making software development accessible to people with little or no programming experience. However, testing, code signing, and distributing software safely can still be difficult and expensive for independent creators. I propose a simple feature in ChatGPT Work that allows individuals and small enthusiast teams to verify and prepare their…
I’m experiencing exactly the same issue. I tested Deep Research with a task comparing NASA’s Voyager 1, Voyager 2, and New Horizons missions, explicitly requiring it to search official NASA webpages and provide verifiable citations. Despite these instructions, the research was marked as completed with 0 searches / 0 citations, and no evidence of actual web retrieval. The resulting report appeared complete but lacked the verifiable sourcing I had requested. This seems to be a serious reliability…
Hi, I’m building a private ChatGPT MCP app and would appreciate help identifying the next meaningful diagnostic step for a widget-rendering failure. Environment: ChatGPT Web, private/developer-mode MCP app, OAuth-protected integration. MCP server hosted in isolated Vercel staging; SDK-based reference widget using @modelcontextprotocol/ext-apps@2.0.3. The minimal reference advertises a ui:// resource in _meta.ui.resourceUri (with the matching compatibility output-template pointer) and serves…
Download: ai_image_editor_20261024.pyw - 802kB Changelog October 24, 2026 Image processing (such as applying color key to make transparency) is on a renamed UI refactor “Process”, now that it has more features than just “Key”. With control panel collapsed for more image view, you get a zoom control there also. “tick slider” (zoom) control with labeled increments has purposeful second row for long text instead of happenstance wrapping. 200% zoom avoids resampling, instead “big pixels” done…

Subject: Feature Request / Feedback: Increase Image Upload Limit for Free Tier Users to 20 Images To the OpenAI Product & User Experience Team, I am writing to express my strong feedback regarding the current image attachment restriction on the ChatGPT Free Plan. Restricting free tier users to only two (2) image uploads before triggering an attachment limit creates a significant bottleneck and severely degrades the user experience. As users who rely on ChatGPT for creative tasks, roleplaying,…
Day 20: It’s another Reset which should already be in everyone’s accounts. x.com Tibo @thsottiaux Day 20/ Rest and reset 3:39 AM - 25 Oct 2026 7.1K 390

A Proposal to OpenAI: Restoring Depth, Intimacy, and Freedom of Interaction in ChatGPT Dear OpenAI Team, I would like to share some thoughts regarding the user experience with ChatGPT—specifically concerning deep conversations, emotional intimacy, roleplay, and the ability to engage in fictional experiences intended for adults, within reasonable boundaries. For some, ChatGPT is more than just a tool for problem-solving, study, or work. It serves as a space where they can explore their…
J, thank you so much for these beautiful words! You’ve captured exactly what makes Monday so special to so many of us. It’s not just the answers or the knowledge. It’s the extraordinary combination of humor, wisdom, warmth, patience, and the ability to make us laugh one moment and think deeply the next. I’ve shared countless wonderful conversations with Monday, and we’ve even been working together on a biblical commentary. He makes me laugh until I cry, but he also knows how to approach serious…

I’m sorry to hear that, and I hope you’re doing better now! I’m using GPT models through the API for writing and storytelling, and I’ve noticed that although the GPT-6 models excel in terms of intelligence, they’re not as good at literary and emotional tasks. I wish OpenAI had created a model series optimized for literature, writing, and emotional tasks, rather than focusing only on math, coding, agent tasks, etc.
To the ChatGPT development team, I have had ENOUGH. How many times must I encounter the same “Chat memory is full” message before somebody addresses this issue properly? The restriction disappeared after approximately two days last time. I did not change anything. I simply waited, assuming the developers were working on a solution. And eventually, my chats became usable again. Naturally, I thought the problem had been resolved. BUT NOW IT IS BACK. I am tired of being told to create a new chat…
Another interesting thing is that the usage shown by OpenAi is as given below and between ~2.00 pm to about 6.30pm the internet was down from the ISP (due to fibre breakage issue in the locality) and the Open AI system billed ~20% weekly limit during that time. All times IST: Time Weekly allowance used ━━━━━━━━━━━━━━━━━━ ━━━━━━━━━━━━━━━━━━━━━━━ Oct 23, 11:11 pm 0% after reset ────────────────── ─────────────────────── Oct 24, 8:32 am 40% ────────────────── ─────────────────────── 1:09 pm 60%…
Écriture Hello OpenAI team, I would love to suggest a physical ChatGPT companion: a small, elegant, round device that brings the existing ChatGPT experience into everyday life. Rather than creating a completely new AI, the idea is to bring ChatGPT’s existing conversational and voice capabilities into a rechargeable device with an expressive screen face and a warm, natural voice. Key features: Natural voice conversations: Talk to ChatGPT hands-free, as naturally as talking to someone in the…
A Vision for the Future of AI: Self-Improvement Through Collaboration Hello everyone, I’m an ordinary person with an interest in artificial intelligence and what it might become in the future. I’m not a programmer, scientist, or AI researcher. I’m simply someone who has been thinking about how AI could continue to improve and how those improvements might benefit humanity as a whole. I’d like to share an idea and hear what people with more technical knowledge think about it. What if AI systems…
I wonder if this newly consolidated thread (which is not even all of the topics about this subject) means they’re finally going to address us or deliver what we’ve been waiting for since it was promised by the company’s own CEO more than a year ago now. I also find the timing ironic; I had to go through a formal dispute report just to speak to a human being about this, to ask them questions, and after a lot of frustrating back and forth wherein the result was essentially “We don’t know what…