ChatGPT Context Limit Reached: What To Do (2026)

ChatGPT is telling some Free users that the “Context window full” or “Chat memory full” notice has been reached on a conversation.

The notice is about one long chat, not your whole account, and a new chat is free to start straight away.

Everything below comes from OpenAI’s own help pages, read on November 20, 2026, and where OpenAI gives no number this guide says so.

Quick answer: what the message means

QuestionAnswer (OpenAI help, checked November 20, 2026)
What is full?The conversation you have open. Your messages and ChatGPT’s replies both use up its space.
Is my account blocked?No. OpenAI says the limit applies only to that conversation.
Fastest fixStart a new chat. OpenAI says you can do that right away for free.
Do I need to clear memory?No. OpenAI says deleting saved memories does not free up room in the chat.
How big is the context window?OpenAI’s help page does not give a size for any plan.

What “Context window full” actually means

A context window is how much information ChatGPT can use in the current conversation at once.

OpenAI says long exchanges use more room than short ones, so there is no fixed number of messages per chat.

The notice can appear the moment you reopen an old chat, even if you never saw a warning while writing in it.

What the Free plan allows

OpenAI’s Free Tier FAQ says Free users have unlimited everyday text chats, subject to abuse-prevention safeguards.

File uploads, image generation, voice, data analysis and other tools have their own separate limits, and the page gives no numbers for them.

ItemWhat OpenAI saysNumber given?
Everyday text chatsUnlimited, subject to safeguardsNo cap stated
Uploads, images, voice, data analysisSeparate usage limitsNo
Messages per chatNo fixed number; depends on lengthNo
Library storage500 MBYes, 500 MB
Default modelThe page lists GPT-5.6 Luna for Free and says it can change over timeNot a limit

What to do when you see the notice

  1. Start a new chat. If the notice offers a summary, you can use it to carry on, but OpenAI warns the summary may leave out details.
  2. Reopen the full chat from your history and copy anything you still need before moving on.
  3. If the notice offers an upgrade, that option lets you continue in the original chat.
  4. If the new chat made from a summary also shows the notice, choose New chat in the sidebar and send a short, separate message.

If a brand new chat says it is full after a few short messages

OpenAI says this can be caused by an ongoing ChatGPT incident or an account issue.

  1. Check OpenAI Status for an incident affecting conversations and try again after it is resolved.
  2. If there is no incident, sign in to the same account in a private browser window and send a short message in a new chat.
  3. If it still happens, contact OpenAI Support.

What to tell support

OpenAI asks you to include your app or browser version, and the time and time zone it happened.

  • Roughly how many messages you sent.
  • What happened when you tried the private browser window.
  • A link to a test chat, if you are comfortable sharing one; your original conversation can stay private.

How to make long chats last longer

OpenAI’s page does not publish tips for stretching a chat, so these are simple habits that follow from how it describes the limit.

  • Open a new chat for each separate task instead of one endless thread.
  • Keep a short note of the facts you want ChatGPT to remember and paste it into the new chat.
  • Copy the important output out of old chats before you start fresh.

OpenAI says the notice may offer an upgrade to continue in the original chat, and that Free usage rates reset when you move to a Plus plan.

Prices and plan limits change, so check the current figures in our ChatGPT Free vs Go vs Plus vs Pro guide or on OpenAI’s pricing page; this post does not quote a price OpenAI’s help pages did not give.

Context full versus a usage limit: which notice is it?

Two different messages look alike, and they need different handling.

NoticeWhat it meansWhen it clears
Context window full / Chat memory fullThat one conversation is too longNever for that chat; start a new chat right away
Free-tier model or GPT limitYour allowance for the default model or for GPTs is used upOpenAI says ChatGPT shows the reset time in the product
Image creation limitThe free image limit is reached; the notice includes an upgrade optionOpenAI gives no timing; you wait until a later time
Uploads, voice, data analysisEach tool has its own separate limitOpenAI gives no timing

The first row is the one this guide covers, and it is the only one where a fresh chat solves the problem immediately.

Things that do not fix it

OpenAI states that you do not need to turn off Memory, delete memories or clear your other chats, because the limit belongs to the single conversation.

  • Deleting saved memories frees nothing in the full chat.
  • Deleting other chats does not give the full chat more room.
  • Switching to a new chat does, which is why OpenAI points you there first.

If you are comparing paid tiers before upgrading, our Claude Pro vs Max and Google AI Plus vs Pro vs Ultra guides cover the main alternatives.

What to use instead

If you would rather not restart a long chat, these assistants have a free tier. Each allowance below is from that company’s own page, checked November 20, 2026.

AssistantFree allowanceSource
ClaudeSession limit that resets every five hours; no message count givensupport.claude.com, Get started with Claude
Google GeminiStandard limits, refreshing every five hours until a weekly limit; 32k token context without a plansupport.google.com/gemini
PerplexityPractically unlimited basic searches; 3 Pro Searches a dayperplexity.ai Help Center
Mistral Le ChatLimited messages and web searches; no number statedmistral.ai/pricing
GrokLimited free Chat and Voice; no number stateddocs.x.ai

Sources

OpenAI help: Context window full or Chat memory full and OpenAI help: ChatGPT Free Tier FAQ, both read November 20, 2026.

Leave a Reply

Your email address will not be published. Required fields are marked *