“Unlimited everyday text chats, subject to abuse-prevention safeguards”
Free and Go users can keep typing on GPT-5.6 Luna without a message counter. Only automated or extreme volume triggers a safeguard.
Reviewed guide · September 1, 2026
ChatGPT limits after unlimited text: what still caps you
On 6 August 2026 OpenAI announced unlimited everyday text chats for Free and Go users on GPT-5.6 Luna, rolling out the following week. Paid plans already had no fixed text cap. The limits did not disappear — they moved. This page maps where they live now.
Reviewed: . Figures marked official come from OpenAI’s help center; everything else is labelled as observed or approximate.
The change
What OpenAI changed in August 2026.
Three sentences from OpenAI’s own help article, and what each means for you.
“Separate limits continue to apply to file uploads, image generation, voice, data analysis”
These are the caps people now hit. Each has its own allowance and reset, and the in-app message names which one.
“If you reach a GPT-5.6 reasoning limit, ChatGPT may continue with another available reasoning model”
Reasoning (the Think option, Medium, High, Extra High, Pro) is metered per plan. Hitting it does not stop the chat; it changes the model.
“OpenAI Support does not reset ChatGPT or Codex usage limits”
No ticket will lift a cap. Waiting for the displayed reset, switching model or upgrading are the only levers.
The map
Where the limits are now, per plan.
Official where OpenAI publishes it; “capped” where OpenAI confirms a limit exists but does not publish the number.
| Plan | Price | Text chat | Reasoning options | Images / files / voice | Deep research | Codex |
|---|---|---|---|---|---|---|
| Free | $0 | Unlimited (Luna) | Think only | Capped; 3 file uploads per day | Limited monthly tasks | — |
| Go | $8/mo, select countries | Unlimited (Luna) | Think only | Capped, higher than Free | Limited monthly tasks | — |
| Plus | $20/mo | Unlimited | Instant, Medium, High (Sol) | Capped; 20 files per project | Monthly allowance | Included, token-metered |
| Pro | $100 or $200/mo | Unlimited | Instant through Pro (Sol + Sol Pro) | Highest consumer allowances; 40 files per project | Largest monthly allowance | Included, more headroom |
| Business | Per seat | Sol 10–100, Terra 25–200, Luna 250–2,000 per 5 h (Standard seats, official) | Instant through Pro | Capped; 40 files per project | Monthly allowance | Included; workspace credits |
| Enterprise | Contract | Per contract; Premium seats 5x with no 5-hour limit | Instant through Pro | Per contract | Per contract | Included; credits |
Prices are as listed on chatgpt.com/pricing at review time; Go availability depends on country. The Business figures are OpenAI’s own “local-message estimates” and the only per-window numbers it publishes.
Decoder
Which limit did you actually hit?
The wording in the app tells you which allowance is involved and what fixes it.
| What ChatGPT says | What it is | What to do |
|---|---|---|
| A reasoning limit with a reset time | Reasoning allowance for your plan tier | Keep chatting on the fallback model, or wait for the time shown |
| “You’ve reached the limit for image generation” | Image allowance | Text still works. Wait, or upgrade if you generate daily |
| Upload refused or “3 files per day” | File allowance (Free) or size cap | Paste text instead, split the file, or move to Plus |
| Deep research counter at zero | Monthly research tasks | Resets 30 days after first use; use normal search-enabled chat meanwhile |
| “Unusual activity” or a pause on plain text | Abuse-prevention safeguard | Stop retrying, wait, reduce volume; nothing in your plan changes this |
| “I can’t help with that” | Policy, not usage | Waiting or upgrading changes nothing; rephrase or accept the refusal |
Make it last
Spend limited usage where it matters.
Default to Instant
Plain text is unlimited. Use the default model for drafting, rewriting and questions, and reserve Think or High reasoning for problems that need it.
Batch images
Write one careful image prompt with size, style and content instead of five small corrections.
Paste, don’t upload
On Free, the three-file daily cap is the tightest limit. Pasting text into the chat costs nothing.
Scope deep research
One precise research question beats three vague ones when the counter is monthly.
Read the reset time
It is the one accurate number in the system. Plan around it instead of retrying.
Alternatives
What to use when this limit stops you.
If a cap lands mid-task and you cannot wait for the reset, the practical fix is a second tool. None of these are affiliated with OpenAI; pick whichever fits the job.
Alternative
Claude
claude.aiStrong writing, analysis and long documents. Has its own five-hour and weekly limits.
AlternativeGemini
gemini.google.comBest inside Google Workspace and for Google-context tasks.
ResearchPerplexity
perplexity.aiResearch answers with web sources and citations.
AlternativeCopilot
copilot.microsoft.comUseful inside Microsoft 365 and Windows.
Multi-modelMultipleChat
multiplechat.aiSeveral models in one workspace so one provider’s cap does not end your day.
Disclosure: MultipleChat is built by the team that runs this site. It is listed as one option among several; nothing on this page is paid placement.
Official sources
Where to verify the current rule.
Checked September 1, 2026. OpenAI changes plan rules without notice; these pages and your own in-app counter are authoritative.
OpenAI Help: GPT-5.6 in ChatGPT
Which reasoning options each plan gets, unlimited text on Free and Go, and what happens at a reasoning limit.
ChatGPT pricing
Current plan names, prices and what each tier includes.
OpenAI Help: ChatGPT Business models and limits
The only official per-window message estimates OpenAI publishes (Standard and Premium seats).
OpenAI Help: File uploads FAQ
Per-file size, files per project, rolling upload cap, Free daily cap and storage limits.
OpenAI Help: Deep research FAQ
How deep research allowances are counted and reset.
FAQ
Unlimited text, explained.
Is ChatGPT really unlimited now?
Everyday text chat is, on every plan including Free and Go, since OpenAI’s August 2026 change — subject to abuse-prevention safeguards. Reasoning modes, images, file uploads, voice, data analysis, deep research and Codex still have limits.
Why did I still hit a limit on a Free account?
You most likely used a limited feature: the Think option, an image, a file upload, voice, or you triggered an abuse safeguard by sending very high volumes. The in-app message names the feature and the reset time.
Does Plus still have message limits?
Not for plain text chat. Plus has limits on the Medium and High reasoning options, on images, uploads, voice and deep research, and on Codex usage.
What does “abuse-prevention safeguards” mean?
OpenAI does not publish the threshold. In practice it targets automated or extremely high-volume use, not a person typing normally. If you hit it while working by hand, wait and reduce volume rather than retrying in a loop.
Can OpenAI Support raise or reset my limit?
No. OpenAI’s help center states that Support does not reset ChatGPT or Codex usage limits. Wait for the reset shown, switch model, or move up a plan.
Where do I see my remaining usage?
ChatGPT shows a message with the reset time when a limit is reached, and deep research shows a task counter. There is no single usage dashboard for consumer plans as of September 2026.