--- title: "Find which ChatGPT memory limit you hit." description: "Separate context-window loss, saved-memory failures, and custom-instruction limits. Use the symptom checklist to choose your workaround." canonical: "https://scalewithsearch.com/articles/chatgpt-memory-limit" date: "2026-01-27" modified: "2026-09-25" --- ## Site navigation - [Scale With Search](https://scalewithsearch.com/) - Real estate - Real estate - [Real estate](https://scalewithsearch.com/for/real-estate) - Work - Start here - [Send your brief](https://scalewithsearch.com/work#send-your-brief) - [Prepare your six-question brief](https://scalewithsearch.com/work#prepare-your-six-question-brief) - Build - [Site build, content library with SEO, signal desk](https://scalewithsearch.com/work) - For your business - Trades and home services - [Auto body and collision shops](https://scalewithsearch.com/for/auto-body-and-collision-shops) - [Foundation and home repair contractors](https://scalewithsearch.com/for/foundation-and-home-repair) - [Garage door and fencing contractors](https://scalewithsearch.com/for/garage-door-and-fencing-contractors) - [HVAC contractors](https://scalewithsearch.com/for/hvac-contractors) - [Janitorial and commercial cleaning companies](https://scalewithsearch.com/for/janitorial-and-commercial-cleaning) - [Locksmiths](https://scalewithsearch.com/for/locksmiths) - [Moving companies](https://scalewithsearch.com/for/moving-companies) - [Pest control companies](https://scalewithsearch.com/for/pest-control-companies) - [Plumbing and electrical contractors](https://scalewithsearch.com/for/plumbing-and-electrical-contractors) - [Restoration and water or fire damage companies](https://scalewithsearch.com/for/restoration-and-water-fire-damage) - [Roofing companies](https://scalewithsearch.com/for/roofing-companies) - [Towing companies](https://scalewithsearch.com/for/towing-companies) - [Tree services and landscaping companies](https://scalewithsearch.com/for/tree-services-and-landscaping) - [Solar installers](https://scalewithsearch.com/for/solar-installation) - [General contractors](https://scalewithsearch.com/for/general-contractors-and-construction) - [Paving, concrete, and flooring contractors](https://scalewithsearch.com/for/paving) - Practices and professional services - [Bookkeeping and tax practices](https://scalewithsearch.com/for/bookkeeping-and-tax-practices) - [Dental practices](https://scalewithsearch.com/for/dental-practices) - [Family and criminal defense law firms](https://scalewithsearch.com/for/family-and-criminal-defense-law-firms) - [Med spas and aesthetics practices](https://scalewithsearch.com/for/med-spas-and-aesthetics) - [Personal injury law firms](https://scalewithsearch.com/for/personal-injury-law-firms) - [Veterinary clinics](https://scalewithsearch.com/for/veterinary-clinics) - [Gyms and fitness studios](https://scalewithsearch.com/for/fitness) - [Therapy and outpatient health practices](https://scalewithsearch.com/for/therapy-and-outpatient-health) - [Medical billing companies](https://scalewithsearch.com/for/medical-billing) - [Insurance agencies](https://scalewithsearch.com/for/insurance-agencies) - [Financial advisors](https://scalewithsearch.com/for/financial-advisors) - [Property management companies](https://scalewithsearch.com/for/property-management) - [Recruiting and staffing agencies](https://scalewithsearch.com/for/recruiting-and-staffing) - [Architects and interior designers](https://scalewithsearch.com/for/architects-and-interior-designers) - [Logistics and supply chain companies](https://scalewithsearch.com/for/logistics-and-supply-chain) - Agencies, MSPs, and manufacturing - [IT and managed service providers](https://scalewithsearch.com/for/it-and-managed-service-providers) - [Machine shops and precision manufacturers](https://scalewithsearch.com/for/machine-shops-and-precision-manufacturing) - [Marketing agencies and freelancers](https://scalewithsearch.com/for/marketing-agencies-and-freelancers) - [SEO agencies and consultants](https://scalewithsearch.com/for/seo-agencies-and-consultants) - [Small manufacturers and fabricators](https://scalewithsearch.com/for/small-manufacturers-and-fabricators) - Restaurants, shops, studios, and nonprofits - [Restaurants and hospitality businesses](https://scalewithsearch.com/for/restaurants-and-hospitality) - [Retail stores and ecommerce sellers](https://scalewithsearch.com/for/retail-and-ecommerce) - [Photographers, event planners, and travel agents](https://scalewithsearch.com/for/photographers) - [Churches and nonprofits](https://scalewithsearch.com/for/churches-and-nonprofits) - [All industries](https://scalewithsearch.com/for/) - Learn - For your office - [Office job guides](https://scalewithsearch.com/guides/) - [Browser calculators](https://scalewithsearch.com/tools/) - Start here - [How it works](https://scalewithsearch.com/how-it-works) - [Free Starter Kit](https://scalewithsearch.com/kit/business-memory-starter-kit.zip) - [Synthetic specimen](https://scalewithsearch.com/specimen/working-session-specimen.zip) - Guides - [The Complete Guide to Business Memory for AI Agents](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide) - [The Complete Small-Business Guide to AI Agent Governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business) - [The Complete Guide to Leaving Vendor AI Memory](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide) - Articles by cluster - [Business memory](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide) - [Agent governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business) - [Migration and ownership](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide) - For machines - [llms.txt](https://scalewithsearch.com/llms.txt) - [llms-full.txt](https://scalewithsearch.com/llms-full.txt) - [Machine view](https://scalewithsearch.com/?view=machine) - Company - Evidence - [Proof](https://scalewithsearch.com/proof) - Company - [About](https://scalewithsearch.com/about) # Identify which ChatGPT memory limit you hit and work around each one. "ChatGPT's memory limit" usually names three different limits at once. One is the context window, which holds the current conversation. One is the saved-memory feature, which keeps short facts between conversations. One is custom instructions, which carry standing guidance into every chat. Each limit fails in its own way, and each has its own workaround. If you fix the wrong one, the problem stays. This page separates them and gives you a quick way to tell which one you hit. It also shows where a context file replaces all three for business work. ## Limit 1: the context window holds one conversation The context window is the maximum amount of text the model can read at one time. It holds your messages, the model's replies, uploaded files, and any instructions in force. It is measured in tokens; one token averages about three quarters of an English word. ChatGPT's window depends on your plan and model, and OpenAI changes the figures with new releases. Check the current plan page for the number that applies to you. For scale, a window of 128,000 tokens holds roughly 96,000 words. Even 96,000 words sounds like a lot. A detailed working session on a complex project, with pasted documents and long replies, can use it up within an hour. Past the limit, the model "forgets" the earliest parts of the conversation. The messages are still visible in the chat, but the model can no longer read them when it writes a reply. You can state something important at the start of a long session, and later the model is unable to see it. The window is a hard ceiling. The [guide to ChatGPT forgetting in the same chat](/articles/why-chatgpt-forgets-same-chat) shows how to test that case. ## Limit 2: saved memory holds a few simple facts OpenAI's memory feature stores facts about you across conversations: your name, your job, basic preferences. Its limits are space and scope. The store has a capacity, but OpenAI does not publish a fixed count. Each saved fact is a short, simple statement. The feature cannot hold complex context such as your business model, your client relationships, or the decisions you made over time and why. Saved memory is also selective. A saved fact applies when the model judges it relevant, not in every chat. The [guide to ChatGPT memory that does not apply](/articles/chatgpt-memory-not-applied-every-conversation) tests that behavior across five kinds of chat. Think of saved memory as a sticky note on the model's desk. It works for a quick reminder. It does not work for deep context. ## Limit 3: custom instructions hold a paragraph of guidance Custom instructions define how ChatGPT should behave in every ordinary conversation. They persist, but they are short. In January 2026, the field held about 1,500 characters, roughly 250 words. As checked on September 19, 2026, OpenAI lists 1,500 characters for Free and Go accounts and 5,000 for paid and education plans. Even the larger figure holds basic preferences, not the context of a business with any complexity. ## Compare what each limit can hold | Limit | Capacity, as observed | What it can hold | Where it applies | |---|---|---|---| | Context window | Set by plan and model; check the current plan page | The current conversation | This chat only | | Saved memory | Short, simple facts; no published count | Short statements | Any chat, when judged relevant | | Custom instructions | 1,500 or 5,000 characters by plan (September 2026) | Basic preferences | Every ordinary chat | None of the three holds what business work needs: complete, current context about your frameworks, client details, and past decisions. That record is what makes an assistant useful, and it has no home in any of the three. ## Tell which limit you hit Use the symptom to find the limit before you choose a workaround. 1. If the model lost a detail from earlier in the same long chat, you hit the context window. 2. If a new chat lacks a fact that you never saved, no limit failed; the fact lived only in the old chat. 3. If a new chat lacks a fact that appears in your saved-memory list, the relevance step skipped it. 4. If the saved-memory list lacks a fact that ChatGPT confirmed it would keep, the save failed or the store was full. 5. If a standing rule is missing because the field ran out of space, you hit the custom instructions limit. Check the saved-memory list in Settings, under Personalization, for cases 3 and 4. Record the date and plan, because labels and behavior change between releases. ## Know why the limits stay It is tempting to expect a fix: a bigger window, more saved facts, longer instructions. OpenAI raised the custom instructions limit for paid plans in July 2026, but each limit still has a cost behind it. A larger context window costs more to process. Standard transformer attention grows with the square of the input length, so a longer window costs much more compute to serve. Saved memories need curation to avoid conflicts; more of them means more contradictions to resolve. Custom instructions apply to every response, so a longer field adds text to every request and can slow each reply. These are technical and economic limits. They move with each model generation, but they do not disappear. ## Work around the context window For long sessions, manage the window directly: - Start a new conversation for a new topic. - Keep related work in the same thread, and unrelated work out of it. - When a thread gets long, ask for a summary of the decisions so far. - Start a fresh chat with that summary as the first message. These steps help, but each one takes effort from you, and none of them carries context into next week. ## Replace memory and instructions with a context file For the saved-memory and custom-instructions limits, the durable fix is a context file outside the chat product. You write your business details, preferences, and operating knowledge in Markdown files. An agent reads them at the start of each session. Claude Code supports this directly. It reads a `CLAUDE.md` file from the project folder at the start of each session. It can open other files in the folder when a task needs them. If the folder is an Obsidian vault, the file can point to the rest of your notes. Your practical memory limit becomes your disk, not a product field. The [guide to context windows and persistent memory](/articles/context-window-vs-persistent-ai-memory) explains how the two work together. The file loads into the window at session start, so a lean file leaves more room for the work. A context file differs from custom instructions in size and form. Custom instructions are a sticky note. A context file can be an employee handbook. It can hold business operations, the client roster, decision frameworks, and communication style, in as much detail as the model's window allows. The file stays on your disk, and Git can record every change. Claude Code still sends the file's contents to the model in each session, so your plan's data terms apply to them. Keep passwords, keys, and regulated records out of the file. ## Check whether ChatGPT's own features are enough ChatGPT's features are enough when your needs are simple: a few personal preferences, short tasks, and conversations that do not depend on last week. Past that point, the limits compound. You manage context by hand, you cannot tell whether the model still has the important detail, and each new chat starts behind. A maintained context file removes most of that overhead. The model starts each session with your current record and refers to past decisions without an explanation. Its first draft is closer to your voice and standards. The [guide to making ChatGPT remember previous conversations](/articles/make-chatgpt-remember-previous-conversations) shows how to move the facts that matter out of old chats and into that record. ## Related: AI memory - [Fix the context, not the prompt, so ChatGPT gives consistent answers](/articles/chatgpt-gives-different-answers) - [Trace what ChatGPT forgot and store it where it stays](/articles/chatgpt-keeps-forgetting-what-i-told-it) - [ChatGPT Project Instructions vs Custom Instructions: Published Character Limits and Scope](/articles/chatgpt-custom-vs-project-instructions-limits) ---- ```text .|########||. .|########||. .|########||. |##||. .||##|. |##||. .||##|. |##||. .||##|. |#|. .|#|. |#|. .|#|. |#|. .|#|. |#| |#| |#| |#| |#| |#| .#| |#. .#| |#. .#| |#. .#. .#| .#. .#| .#. .#| |#. .#| |#. .#| |#. .#| |# ...... #| |# ...... #| |# ...... #| .# ||#########| #| .# ||#########| #| .# ||#########| #| .#. |######||######|. .#. .#. |######||######|. .#. .#. |######||######|. .#. #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# # .||####|###########################|. # # .||####|###########################|. # # .||####|###########################|. # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || #. .######|#|#########| .# #. .######|#|#########| .# #. .######|#|#########| .# .#. |#####||||#####| .#. .#. |#####||||#####| .#. .#. |#####||||#####| .#. |# ||########|| #. |# ||########|| #. |# ||########|| #. |# ...... #| |# ...... #| |# ...... #| |#. .#| |#. .#| |#. .#| |#. .#. |#. .#. |#. .#. .#| |#. .#| |#. .#| |#. |#| |#| |#| |#| |#| |#| .|#|. .|#|. .|#|. .|#|. .|#|. .|#|. |##||. .||##| |##||. .||##| |##||. .||##| .||########||. .||########||. .||########||. Scale With Search 2026 [scalewithsearch.com](https://scalewithsearch.com) ```