# \#memory **URL:** https://community.openai.com/tag/memory/595.md [Latest](https://community.openai.com/latest.md) · [Categories](https://community.openai.com/categories.md) · [Tags](https://community.openai.com/tags.md) --- ## [Keep the Why: Git-native project memory for coding agents](https://community.openai.com/t/keep-the-why-git-native-project-memory-for-coding-agents/1404573)
**Author:** [@oliver-zehentleitner](https://community.openai.com/u/oliver-zehentleitner)\ **Replies:** 10\ **Last updated:** [October 18, 2026, 8:44pm UTC](https://community.openai.com/t/keep-the-why-git-native-project-memory-for-coding-agents/1404573 "2026-10-18T20:44:05Z")
I’ve been working on an open source project called Keep the Why. The problem I wanted to solve is pretty simple: Coding agents are getting very good at understanding code, but across sessions they often lose the reason… --- ## [Persistent AI Characters, Verifiable Memory, and Long-Running Project Continuity](https://community.openai.com/t/persistent-ai-characters-verifiable-memory-and-long-running-project-continuity/1405733)
**Author:** [@fubunker](https://community.openai.com/u/fubunker)\ **Replies:** 15\ **Last updated:** [October 18, 2026, 1:54pm UTC](https://community.openai.com/t/persistent-ai-characters-verifiable-memory-and-long-running-project-continuity/1405733 "2026-10-18T13:54:47Z")
What we’re building I’m developing FakeAwakeAI, an experimental creative universe containing eleven distinct AI characters, each with its own personality, relationships, responsibilities, history, and perspective. For … --- ## [What if ChatGPT Could Become the ONE Personal AI Companion You Actually Live With?](https://community.openai.com/t/what-if-chatgpt-could-become-the-one-personal-ai-companion-you-actually-live-with/1404521)
**Author:** [@Dreamwhopper](https://community.openai.com/u/Dreamwhopper)\ **Replies:** 17\ **Last updated:** [October 18, 2026, 3:37am UTC](https://community.openai.com/t/what-if-chatgpt-could-become-the-one-personal-ai-companion-you-actually-live-with/1404521 "2026-10-18T03:37:11Z")
Feature Request: ChatGPT Living Avatar — Why Have Five Devices When One AI Could Do It All? I can choose ChatGPT’s voice. Why can’t I choose the face that goes with it? That simple question led me to something much big… --- ## [ChatGPT Year in Review (“ChatGPT Wrapped”)](https://community.openai.com/t/chatgpt-year-in-review-chatgpt-wrapped/1406816)
**Author:** [@lostremains](https://community.openai.com/u/lostremains)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 5:15am UTC](https://community.openai.com/t/chatgpt-year-in-review-chatgpt-wrapped/1406816 "2026-10-16T05:15:47Z")
Feature Request: ChatGPT Year in Review (“ChatGPT Wrapped”) I recently asked ChatGPT a simple question: “How many conversations have we had?” ChatGPT explained that although I can see my individual conversations in the… --- ## ["Chat memory full". Can you communicate with users regarding new usage limits imposed on free plans](https://community.openai.com/t/chat-memory-full-can-you-communicate-with-users-regarding-new-usage-limits-imposed-on-free-plans/1406722)
**Author:** [@Louis25](https://community.openai.com/u/Louis25)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 8:12pm UTC](https://community.openai.com/t/chat-memory-full-can-you-communicate-with-users-regarding-new-usage-limits-imposed-on-free-plans/1406722 "2026-10-15T20:12:01Z")
We all know what I’m talking about. If this is indeed not a bug, it is clearly a new usage limit. We received no communication on this before it happened, and nothing is being said currently, many weeks in. I understand… --- ## [Future product idea: ChatGPT as a persistent humanoid companion](https://community.openai.com/t/future-product-idea-chatgpt-as-a-persistent-humanoid-companion/1404064)
**Author:** [@roaldkauan](https://community.openai.com/u/roaldkauan)\ **Replies:** 1\ **Last updated:** [October 15, 2026, 9:26am UTC](https://community.openai.com/t/future-product-idea-chatgpt-as-a-persistent-humanoid-companion/1404064 "2026-10-15T09:26:22Z")
Future Product Concept: A Persistent ChatGPT Humanoid Companion I would like to propose a future OpenAI product concept: a humanoid robot that could serve as a physical extension of a user’s ChatGPT assistant. The key … --- ## [Saved Memory becomes unavailable after using a connected app — reproducible within a session, and not limited to Google connectors](https://community.openai.com/t/saved-memory-becomes-unavailable-after-using-a-connected-app-reproducible-within-a-session-and-not-limited-to-google-connectors/1395159)
**Author:** [@okanotamo](https://community.openai.com/u/okanotamo)\ **Replies:** 14\ **Last updated:** [October 14, 2026, 10:28pm UTC](https://community.openai.com/t/saved-memory-becomes-unavailable-after-using-a-connected-app-reproducible-within-a-session-and-not-limited-to-google-connectors/1395159 "2026-10-14T22:28:58Z")
Hi everyone, I’ve been running a series of tests on ChatGPT’s Saved Memory and connected apps, and I’d like to share what I’ve found and ask whether others can reproduce it. What happens After a connected app is used i… --- ## [Saved Memory & Improved Memory](https://community.openai.com/t/saved-memory-improved-memory/1406314)
**Author:** [@YoMama8741](https://community.openai.com/u/YoMama8741)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 2:43pm UTC](https://community.openai.com/t/saved-memory-improved-memory/1406314 "2026-10-14T14:43:02Z")
Hi, I have been using ChatGPT for my projects & I would like the saved memory to be compatible with the improved memory feature as when I added some stuff to my saved memory, it immeadiately forgets what I saved memory. … --- ## [Saved Memory not working & co existing with memory summary](https://community.openai.com/t/saved-memory-not-working-co-existing-with-memory-summary/1406009)
**Author:** [@YoMama8741](https://community.openai.com/u/YoMama8741)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 2:25pm UTC](https://community.openai.com/t/saved-memory-not-working-co-existing-with-memory-summary/1406009 "2026-10-13T14:25:46Z")
Hey, I would like to suggest for the legacy memory & the saved memory to co-exist. Sometimes ChatGPT forgets stuff when I put things into memory. I find it annoying having to both toggle on & off. I would like this bug t… --- ## [Feature Request: A Small Physical ChatGPT Companion — Not an AI Pet](https://community.openai.com/t/feature-request-a-small-physical-chatgpt-companion-not-an-ai-pet/1400276)
**Author:** [@heatherholinka](https://community.openai.com/u/heatherholinka)\ **Replies:** 1\ **Last updated:** [October 13, 2026, 1:04pm UTC](https://community.openai.com/t/feature-request-a-small-physical-chatgpt-companion-not-an-ai-pet/1400276 "2026-10-13T13:04:57Z")
I would love to see OpenAI explore a small, affordable physical companion device for ChatGPT. I’m not looking for an AI pet, a virtual puppy, or a robot with its own preprogrammed personality. There are already products… --- ## [Solving agent system prompt drift in long sessions — a 300-token fix](https://community.openai.com/t/solving-agent-system-prompt-drift-in-long-sessions-a-300-token-fix/1375139)
**Author:** [@Nikolasi](https://community.openai.com/u/Nikolasi)\ **Replies:** 3\ **Last updated:** [October 13, 2026, 11:30am UTC](https://community.openai.com/t/solving-agent-system-prompt-drift-in-long-sessions-a-300-token-fix/1375139 "2026-10-13T11:30:30Z")
The problem If you’ve run any LLM agent for 30+ minutes, you’ve seen this: the agent follows its system prompt perfectly at the start, then gradually drifts. An hour in — it acts like the prompt never existed. This hap… --- ## [User Knowledge Memories reverting to older state and mixing outdated context](https://community.openai.com/t/user-knowledge-memories-reverting-to-older-state-and-mixing-outdated-context/1405494)
**Author:** [@qaotic](https://community.openai.com/u/qaotic)\ **Replies:** 1\ **Last updated:** [October 13, 2026, 4:08am UTC](https://community.openai.com/t/user-knowledge-memories-reverting-to-older-state-and-mixing-outdated-context/1405494 "2026-10-13T04:08:20Z")
I may have found an issue where the memory information provided to ChatGPT has reverted to a significantly older state, while older context also appears to be getting mixed into newly formed memories. In April 2026, I a… --- ## [Feature Request — Persistent File Modification Permissions for ChatGPT Library](https://community.openai.com/t/feature-request-persistent-file-modification-permissions-for-chatgpt-library/1405822)
**Author:** [@nicholaselazzaro](https://community.openai.com/u/nicholaselazzaro)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 1:17am UTC](https://community.openai.com/t/feature-request-persistent-file-modification-permissions-for-chatgpt-library/1405822 "2026-10-13T01:17:34Z")
I would like the ability to grant ChatGPT persistent authorization to create, modify, replace, and maintain files within specific folders in my ChatGPT Library, without requiring individual approval for every operation. … --- ## [AI as a tool, an automation platform, or a long-term collaborator — are these becoming different products?](https://community.openai.com/t/ai-as-a-tool-an-automation-platform-or-a-long-term-collaborator-are-these-becoming-different-products/1403250)
**Author:** [@Daareek](https://community.openai.com/u/Daareek)\ **Replies:** 10\ **Last updated:** [October 10, 2026, 5:59pm UTC](https://community.openai.com/t/ai-as-a-tool-an-automation-platform-or-a-long-term-collaborator-are-these-becoming-different-products/1403250 "2026-10-10T17:59:19Z")
I’ve been wondering whether we are starting to talk about “AI users” as if they were one group, while in reality very different usage models are emerging. The more I use AI, the more I see at least three very different … --- ## [ChatGPT recalls saved memory but fails to apply it, and invents plausible context instead of verifying history](https://community.openai.com/t/chatgpt-recalls-saved-memory-but-fails-to-apply-it-and-invents-plausible-context-instead-of-verifying-history/1393739)
**Author:** [@karappo](https://community.openai.com/u/karappo)\ **Replies:** 7\ **Last updated:** [October 9, 2026, 2:18am UTC](https://community.openai.com/t/chatgpt-recalls-saved-memory-but-fails-to-apply-it-and-invents-plausible-context-instead-of-verifying-history/1393739 "2026-10-09T02:18:55Z")
I am reporting what appears to be two related ChatGPT reliability issues: 1. Saved memory/instructions can be recalled, but are not reliably applied during answer generation. 2. When previous conversational context has… --- ## [\[Feature Concept\] What If ChatGPT Could Be Your Professional Reference?](https://community.openai.com/t/feature-concept-what-if-chatgpt-could-be-your-professional-reference/1404429)
**Author:** [@naka98](https://community.openai.com/u/naka98)\ **Replies:** 0\ **Last updated:** [October 8, 2026, 4:32pm UTC](https://community.openai.com/t/feature-concept-what-if-chatgpt-could-be-your-professional-reference/1404429 "2026-10-08T16:32:26Z")
A User-Controlled AI Cognitive Portfolio The Paradigm Shift As AI takes over more technical execution, how people think, make decisions, solve problems, and evolve may become more important than what they personally bu… --- ## [Feature Request: Let ChatGPT Reliably Retrieve Exact Information Across Past Chats](https://community.openai.com/t/feature-request-let-chatgpt-reliably-retrieve-exact-information-across-past-chats/1404280)
**Author:** [@lavskiy.o.v](https://community.openai.com/u/lavskiy.o.v)\ **Replies:** 0\ **Last updated:** [October 8, 2026, 8:21am UTC](https://community.openai.com/t/feature-request-let-chatgpt-reliably-retrieve-exact-information-across-past-chats/1404280 "2026-10-08T08:21:51Z")
ChatGPT already has useful features such as Projects and chat-history search, but I still think there is an important missing capability: reliable assistant-driven retrieval of exact information from previous conversatio… --- ## [Beyond Long-Term AI Memory: Building Persistent, Governed Canonical State with ChatGPT + GitHub + Drive](https://community.openai.com/t/beyond-long-term-ai-memory-building-persistent-governed-canonical-state-with-chatgpt-github-drive/1403168)
**Author:** [@ChipCAD](https://community.openai.com/u/ChipCAD)\ **Replies:** 3\ **Last updated:** [October 7, 2026, 10:59pm UTC](https://community.openai.com/t/beyond-long-term-ai-memory-building-persistent-governed-canonical-state-with-chatgpt-github-drive/1403168 "2026-10-07T22:59:00Z")
I originally thought I was trying to solve long-term memory. What I ended up building was something different: an external canonical-state system with rules governing what can become durable, which sources have authority… --- ## [GPT-5.6 Regression: Memory Retrieval Fails, Long Chats Lose Context Early, and In-Chat Model Switching Is Gone](https://community.openai.com/t/gpt-5-6-regression-memory-retrieval-fails-long-chats-lose-context-early-and-in-chat-model-switching-is-gone/1403213)
**Author:** [@schneider](https://community.openai.com/u/schneider)\ **Replies:** 2\ **Last updated:** [October 5, 2026, 3:47am UTC](https://community.openai.com/t/gpt-5-6-regression-memory-retrieval-fails-long-chats-lose-context-early-and-in-chat-model-switching-is-gone/1403213 "2026-10-05T03:47:37Z")
This appears to be a reproducible product regression affecting longitudinal ChatGPT use on a paid account, not an isolated memory failure. TITLE: Regression in ChatGPT: persistent memory is available but not applied, l… --- ## [Optional Speaker Recognition for ChatGPT Voice](https://community.openai.com/t/optional-speaker-recognition-for-chatgpt-voice/1402974)
**Author:** [@neo82575](https://community.openai.com/u/neo82575)\ **Replies:** 0\ **Last updated:** [October 2, 2026, 9:06pm UTC](https://community.openai.com/t/optional-speaker-recognition-for-chatgpt-voice/1402974 "2026-10-02T21:06:46Z")
ChatGPT Voice can understand what is being said, but it does not currently distinguish which person is speaking during a shared or multi-person conversation. I’d like to propose an optional, opt-in speaker-recognition f… --- ## [ChatGPT Spatial Companion: A Mixed-Reality Headset With Cross-Device Continuity](https://community.openai.com/t/chatgpt-spatial-companion-a-mixed-reality-headset-with-cross-device-continuity/1398127)
**Author:** [@Mr\_Rainbow1982](https://community.openai.com/u/Mr_Rainbow1982)\ **Replies:** 5\ **Last updated:** [October 2, 2026, 9:02pm UTC](https://community.openai.com/t/chatgpt-spatial-companion-a-mixed-reality-headset-with-cross-device-continuity/1398127 "2026-10-02T21:02:05Z")
To Sam Altman and the OpenAI hardware and product teams: Imagine putting on an OpenAI mixed-reality headset and seeing the same personal ChatGPT you have already spoken with for months or years appear life-sized in your… --- ## [Feature proposal: Combine context compaction, procedural memory and continual task learning for long-running agents](https://community.openai.com/t/feature-proposal-combine-context-compaction-procedural-memory-and-continual-task-learning-for-long-running-agents/1402687)
**Author:** [@yewheng86](https://community.openai.com/u/yewheng86)\ **Replies:** 0\ **Last updated:** [October 1, 2026, 8:31pm UTC](https://community.openai.com/t/feature-proposal-combine-context-compaction-procedural-memory-and-continual-task-learning-for-long-running-agents/1402687 "2026-10-01T20:31:14Z")
I recently contacted OpenAI Support with an idea about combining several AI capabilities into a single learning loop for long-running agents. Support mentioned that some of the building blocks already exist, including se… --- ## [Feature Request: Live Context Inheritance Between ChatGPT Projects](https://community.openai.com/t/feature-request-live-context-inheritance-between-chatgpt-projects/1401835)
**Author:** [@Emaidiong](https://community.openai.com/u/Emaidiong)\ **Replies:** 2\ **Last updated:** [September 30, 2026, 11:41am UTC](https://community.openai.com/t/feature-request-live-context-inheritance-between-chatgpt-projects/1401835 "2026-09-30T11:41:08Z")
Summary Allow one ChatGPT Project to use another Project as a live context source, instead of requiring static file imports or manual copy-pasting. For example, I may spend weeks building strategy, brand voice, framewor… --- ## [Request to Remove or Expand the Per-Chat Memory / Context Limit for Free Version](https://community.openai.com/t/request-to-remove-or-expand-the-per-chat-memory-context-limit-for-free-version/1402108)
**Author:** [@AyaanBilal](https://community.openai.com/u/AyaanBilal)\ **Replies:** 1\ **Last updated:** [September 30, 2026, 4:39am UTC](https://community.openai.com/t/request-to-remove-or-expand-the-per-chat-memory-context-limit-for-free-version/1402108 "2026-09-30T04:39:44Z")
The Problem: When engaging in long, complex, or ongoing chat sessions on the Free Version of ChatGPT, users eventually hit a strict per-chat memory capacity (context window limit). Once this limit is reached, the conver… --- ## [GPT-5.6 Instant: Regression in Cross-Chat Context Retrieval and Response Quality](https://community.openai.com/t/gpt-5-6-instant-regression-in-cross-chat-context-retrieval-and-response-quality/1401495)
**Author:** [@ContextDrift](https://community.openai.com/u/ContextDrift)\ **Replies:** 1\ **Last updated:** [September 28, 2026, 7:31am UTC](https://community.openai.com/t/gpt-5-6-instant-regression-in-cross-chat-context-retrieval-and-response-quality/1401495 "2026-09-28T07:31:15Z")
Since the recent GPT-6 Sol/Luna rollout, I have noticed a significant regression when using GPT-5.6 Instant. I am currently subscribed to the Pro 5x plan for context. Context from previous conversations is often missing … --- ## [Goldfish Memory - GPT-5.6: Local intelligence is high, but longitudinal intelligence has become unreliable](https://community.openai.com/t/goldfish-memory-gpt-5-6-local-intelligence-is-high-but-longitudinal-intelligence-has-become-unreliable/1401345)
**Author:** [@keylon99](https://community.openai.com/u/keylon99)\ **Replies:** 2\ **Last updated:** [September 28, 2026, 7:20am UTC](https://community.openai.com/t/goldfish-memory-gpt-5-6-local-intelligence-is-high-but-longitudinal-intelligence-has-become-unreliable/1401345 "2026-09-28T07:20:53Z")
Apologies for the length, but these issues are interconnected. I’ve tried to keep this concise without stripping out the evidence needed to show the pattern. I’ve been using ChatGPT heavily for long-running professional… --- ## [I would like a feature to increase ChatGpt's obedience, reducing disinformacion and avoiding the ''wrong path''](https://community.openai.com/t/i-would-like-a-feature-to-increase-chatgpts-obedience-reducing-disinformacion-and-avoiding-the-wrong-path/1401384)
**Author:** [@SalvaWorks](https://community.openai.com/u/SalvaWorks)\ **Replies:** 0\ **Last updated:** [September 27, 2026, 9:09pm UTC](https://community.openai.com/t/i-would-like-a-feature-to-increase-chatgpts-obedience-reducing-disinformacion-and-avoiding-the-wrong-path/1401384 "2026-09-27T21:09:11Z")
I would like to report a recurring issue with ChatGPT that affects different kinds of tasks. To reduce errors, I have saved memories about my preferences and decisions, set up custom instructions, established rules for h… --- ## [Best workflow to turn long ChatGPT chats into clean Custom GPT instructions?](https://community.openai.com/t/best-workflow-to-turn-long-chatgpt-chats-into-clean-custom-gpt-instructions/1383667)
**Author:** [@Edenlytics](https://community.openai.com/u/Edenlytics)\ **Replies:** 2\ **Last updated:** [September 27, 2026, 7:53pm UTC](https://community.openai.com/t/best-workflow-to-turn-long-chatgpt-chats-into-clean-custom-gpt-instructions/1383667 "2026-09-27T19:53:04Z")
Hi everyone, I’m trying to improve how I use ChatGPT/Custom GPTs for long-term project work. I often have long chats where I build up useful instructions, rules, workflows, and project knowledge over time. The problem i… --- ## [Feature Proposal: Let Users Choose Between Tool Mode and a Proactive Personal Assistant Mode](https://community.openai.com/t/feature-proposal-let-users-choose-between-tool-mode-and-a-proactive-personal-assistant-mode/1401348)
**Author:** [@yuanweiren3](https://community.openai.com/u/yuanweiren3)\ **Replies:** 0\ **Last updated:** [September 27, 2026, 5:51pm UTC](https://community.openai.com/t/feature-proposal-let-users-choose-between-tool-mode-and-a-proactive-personal-assistant-mode/1401348 "2026-09-27T17:51:38Z")
Hi OpenAI team, I’m a long-term ChatGPT user, and I’d like to suggest a broader evolution of ChatGPT from a primarily reactive AI tool into an optional, genuinely proactive personal assistant. My frustration is not tha… --- ## [One-Click "Continue in New Chat" Button for Seamless Thread Migration](https://community.openai.com/t/one-click-continue-in-new-chat-button-for-seamless-thread-migration/1401276)
**Author:** [@neeraj.lilani](https://community.openai.com/u/neeraj.lilani)\ **Replies:** 0\ **Last updated:** [September 27, 2026, 1:40pm UTC](https://community.openai.com/t/one-click-continue-in-new-chat-button-for-seamless-thread-migration/1401276 "2026-09-27T13:40:23Z")
When a chat thread reaches its token limit or enters a state of context degradation, users face a significant workflow disruption. They must manually copy the latest code or project state, open a new chat, and write a pr… [Next page](https://community.openai.com/tag/memory/595.md?match_all_tags=true&page=1&tags%5B%5D=memory)