Troubleshooting

ChatGPT memory not working?

If ChatGPT stopped remembering you — or never started — it's almost always one of a handful of causes: a toggle that's off, memory that's full, a Temporary Chat, the wrong account, or a feature that isn't available to you yet. This guide pairs each cause with its fix and tells you how to verify what's actually happening.

Causes & fixes

Why memory stops working — and what to do

Work down this list in order; the first few catch most cases.

CauseFix
Memory toggle is offGo to Settings → Personalization → Memory and turn on "Reference saved memories" (and chat history if you want it). See turn memory on/off.
Memory is fullOpen "Manage memories" and delete some entries to free space so new facts can be saved. See memory limits.
You're in a Temporary ChatMemory is disabled there by design. Start a normal chat. See Temporary Chat.
Different accountMemory is tied to your account. Confirm you're signed into the same account on every device.
Still rolling out / not on your plan or regionMemory rolls out gradually. Check OpenAI's current availability notes; it may not be available to you yet.
Admin disabled it (Business/Enterprise)On Team/Enterprise, an admin can disable memory. Ask your admin. See memory for business.
It never saved that factMemory is selective. Check "Manage memories" for what's actually stored and add the fact explicitly if needed.
Recall is approximate"Reference chat history" is implicit, so it won't recall exact details. For precision, use saved memories or custom instructions.

Behavior and availability change. Always confirm on OpenAI's Memory FAQ and Data Controls FAQ.

Quick diagnosis

A four-step check

Run these in order before assuming memory is broken.

01

Not a Temporary Chat?

Confirm you're in a normal chat — memory is off in Temporary Chat by design.

02

Right account?

Make sure you're signed into the same account where your memories live.

03

Toggles on?

In Settings → Personalization → Memory, turn "Reference saved memories" (and chat history) on.

04

Room to store?

If memory is full, delete some entries in "Manage memories" so new facts can be saved.

If memory worries you

Use AI without locking all your context to one provider

Memory is convenient, but it concentrates your personal context inside one company's servers. If you'd rather spread your usage — and pick which model sees which prompt — a multi-model workspace gives you that choice while you keep sensitive topics in memory-free sessions.

All-in-one

MultipleChat AI

multiplechat.ai

Use ChatGPT, Claude, Gemini and other models in one place — choose where each prompt goes instead of feeding everything to a single assistant.

Privacy-first: MultipleChat doesn't save your chats to memory and doesn't share your data with model providers or let them train on it. Other privacy-focused options include Duck.ai, Proton Lumo, Brave Leo and Kagi Assistant; for full local control, self-hosted tools like Ollama or Jan keep everything on your device.

OpenAI

ChatGPT

chatgpt.com

Offers Temporary Chat (no history, not used for training), a training opt-out under Settings → Data Controls, and editable memory. By default, consumer chats may be used to improve models unless you opt out — verify on OpenAI's help center.

Anthropic

Claude

claude.ai

Anthropic gives you data and training controls and has historically taken a privacy-conscious stance on consumer chats, but its training and retention settings have changed over time — check your current privacy settings in the app.

Google

Gemini

gemini.google.com

Gemini Apps Activity lets you pause history and delete past activity, but sampled conversations can be reviewed by humans, so avoid sharing confidential information. Review your Activity controls.

Microsoft

Copilot

copilot.microsoft.com

Microsoft 365 Copilot (business) doesn't use your organization's data to train models and adds Enterprise Data Protection; consumer Copilot has a privacy dashboard to review and delete your history. Check current Microsoft policies.

Perplexity

Perplexity

perplexity.ai

Perplexity has an "AI data retention" setting you can switch off so your searches aren't used to improve its models, plus account-level data controls. Confirm the current options in settings.

FAQ

Memory not working — quick answers

The usual causes, how to check each, and the fastest fix.

Why is ChatGPT memory not working?

The most common reasons are: the memory toggle is off, memory is full, you're in a Temporary Chat, you're signed into a different account, the feature is still rolling out or isn't available on your plan or region, a Business or Enterprise admin disabled it, or you asked about something it never saved. Each has a straightforward fix, which you can verify in Settings.

How do I check if memory is turned on?

Go to Settings → Personalization → Memory and confirm "Reference saved memories" (and, if you want it, "Reference chat history") is on. If saved memories is off, ChatGPT won't use stored facts.

What does "memory is full" mean and how do I fix it?

Saved memory has a finite capacity. When it's full, ChatGPT can't store new facts until you make room. Open "Manage memories" and delete some entries to free space. See the memory limits guide for more.

Could Temporary Chat be why memory isn't working?

Yes. Memory is disabled in Temporary Chat by design — it doesn't read or create memories. If memory seems missing, check whether you're in a Temporary Chat and start a normal chat instead.

Why does memory work on one device but not another?

Memory is tied to your account on OpenAI's servers. If it's missing on one device, you may be signed into a different account there. Confirm you're logged into the same account everywhere.

Why don't I have the memory feature at all?

Memory features roll out gradually and may not yet be available on your plan or in your region. On Business and Enterprise plans an admin can disable memory. Check OpenAI's current availability notes and, for managed accounts, ask your admin.

Why doesn't ChatGPT remember something I told it?

It may never have saved that detail as a memory. Saved memories are selective, and "Reference chat history" is implicit, so it won't always recall a specific fact exactly. Check "Manage memories" to see what's actually stored, and add the fact explicitly if needed.

Why does ChatGPT recall some things but not exactly what I expect?

The "Reference chat history" layer is implicit — it draws on patterns from past chats rather than an exact, editable record. So recall can be approximate. For precise control, rely on saved memories or custom instructions instead.

Can a Business or Enterprise admin disable memory?

Yes. On Team and Enterprise plans, admins can control whether memory is available. If you're on a managed account and memory is missing, the admin may have disabled it. See the memory for business guide.

What's the quickest way to fix memory that isn't working?

Confirm you're not in a Temporary Chat, check you're on the right account, open Settings → Personalization → Memory and turn the toggles on, and free space in "Manage memories" if it's full. If memory still isn't there, it may be a rollout or admin restriction — verify on OpenAI's help center.