What is the Claude limit?
On Pro, about 45 messages every 5 hours (roughly 200+ short messages a day), plus a weekly cap. The exact count drops with long chats, big files, Opus/Research and busy periods.
Reviewed guide · July 29, 2026
Hit Claude limit?
Claude limits are not one magic number. Message length, file uploads, conversation length, model choice, Research, Claude Code and current capacity can all change how quickly you hit a cap.
Reviewed: . Estimate basis: official Anthropic pages, in-product usage messages, and observed user reports. Verify your exact account in Settings -> Usage.
Straight answers
What is the Claude limit? The short version.
Anthropic no longer prints one official number, but here are the real figures users are seeing right now (approximate, reviewed July 2026 — check Claude → Settings → Usage for your exact caps). Every Claude plan uses two layers at once: a rolling 5-hour limit and a weekly limit.
| Plan | Price (verify) | Chat messages per 5h (approx.) | Models | Weekly (Claude Code, Sonnet) |
|---|---|---|---|---|
| Free | $0 | A handful to ~15 short messages, highly variable; stops sooner when busy. | Sonnet only | Not for sustained work |
| Pro | ~$20/mo | ~45 short messages per 5 hours (≈200+/day if spread out); ~35–40 at weekday peak (5–11am PT). | Sonnet (chat) | ~40–80 Sonnet hours |
| Max 5x | ~$100/mo | ~5× Pro (≈225 messages per 5h). | + Opus | ~140–280 Sonnet hours |
| Max 20x | ~$200/mo | ~20× Pro (≈900 messages per 5h). | + Opus | ~240–480 Sonnet hours, ~24–40 Opus hours |
What is the Cowork limit?
There is no separate Cowork limit. Claude Cowork draws from the same plan usage pool as Claude.ai chat and Claude Code, so heavy Cowork/agent work eats into your 5-hour and weekly allowance.
What is the Claude Code limit?
Also shared with chat and Cowork. Rough estimates: ~10–45 prompts per 5 hours on Pro, far more on Max 20x (up to ~900), with the weekly Sonnet hours above.
When does it reset?
The 5-hour limit resets ~5 hours after your first message (rolling). The weekly limit resets on a fixed weekly schedule, with separate resets for Opus vs other models — see Settings → Usage.
Alternatives
What to use when Claude is capped.
If Claude stops because of a usage limit, use legitimate alternatives instead of account tricks. MultipleChat is useful when you want Claude-style writing alongside ChatGPT, Gemini, Grok and search-style models in one place, so one app’s cap does not block your whole day; you can keep working on another model and come back to Claude after a reset.
All-in-one
MultipleChat AI
multiplechat.aiUseful when you want Claude, ChatGPT, Gemini, Grok and search-style models in one place. Also helpful when one assistant is capped or not ideal for the task.
AlternativeChatGPT
chatgpt.comBroad assistant for writing, coding, images, files, voice and everyday productivity.
AlternativeGemini
gemini.google.comUseful for Google Workspace and Google-context work.
ResearchPerplexity
perplexity.aiResearch-focused answers with web sources.
WorkMicrosoft Copilot
copilot.microsoft.comPractical for Microsoft users and Office workflows.
Multi-modelPoe
poe.comMulti-bot platform for trying several assistants.
Our guideClaude Alternative
claudealternative.comGuide to Claude alternatives by use case.
ComparisonChatGPT vs Claude
chatsgptvsclaude.comDetailed comparison of Claude and ChatGPT.
GuideChatGPT Alternatives
chatgpt-alternatives.comBroader list of tools similar to ChatGPT.
Limit types
Claude usage limits depend on more than messages.
A short chat and a long document-analysis session do not cost the same. Claude usage is shaped by model, feature, files, context and current capacity. The most useful thing to understand is that there are two layers working together: a rolling five-hour session limit that resets a few hours after your first message, and weekly limits on paid plans with separate resets for Opus and other models. When Claude says you hit a limit, the first job is to work out which layer and which type you actually reached, because the fix is different for each.
Message limits
The visible cap most users notice. It is affected by the model, message length, conversation length and current capacity.
Five-hour sessions
Anthropic says Pro session-based usage limits reset every five hours, but the number of messages varies.
Conversation length
Long threads consume more context. Starting a focused new chat can be more efficient than dragging one chat forever.
File and attachment limits
Big PDFs, long docs, repeated uploads and data-heavy analysis can use more capacity than short text prompts.
Model and feature limits
Opus, Sonnet, Haiku, Research, web search, artifacts, connectors and Claude Code can behave differently.
Output limits
Long generations, code, tables and full-document rewriting can hit output or task limits.
Claude Code limits
Claude Code shares or consumes usage differently from normal chat depending on plan and setup.
Policy restrictions
A refusal can be a safety/policy issue, not a usage cap. Paying more does not remove Anthropic rules.
Search intent
What people mean by “Claude limit”.
Most people who search “Claude limit” have just seen a cap message and do not know which one they hit. These are the different things the same words can mean, so you can fix the real problem instead of guessing a fake daily number.
Daily message cap
The cap most users notice first. There is no single fixed number; it shifts with model, message length, attachments, conversation length and current capacity.
Pro and Max caps
Paid users often expect “unlimited” and are surprised by limits. Pro raises usage over Free; Max raises it further, but neither removes every cap.
Claude Code limits
Developers mix chat caps with agent caps. Claude Code reads files, edits code and runs commands, so it draws on the same usage pool faster than a short chat.
Weekly limit
Paid plans now carry a weekly layer on top of the rolling session limit. Anthropic shows separate weekly resets, including a distinct one for Opus.
Reset time
Many searches are really “when does my limit reset?”. The session limit resets about five hours after your first message; weekly limits reset on their own schedule.
File and PDF limits
People ask whether Claude can read “unlimited” PDFs. Large documents, repeated uploads and heavy extraction use more of your usage than short text.
Output length
Some users hit a wall mid-answer on long rewrites, big tables or large code blocks. That is an output or task limit, not always a usage cap.
Rate limits and API
Developers conflate claude.ai subscriptions with API rate limits. The API has its own token pricing, rate limits and billing, separate from chat.
Policy refusals
A refusal can be a safety or policy decision, not a usage cap. Paying more raises access but does not remove Anthropic’s rules.
Plans
Claude Free, Pro, Max, Team, Enterprise and API limits.
Use this as a planning map, not a fake quota table. There are two layers to keep in mind: a rolling five-hour session limit that resets about five hours after your first message, and weekly limits on paid plans with separate resets for Opus and other models (Settings → Usage). Pricing to verify on the official page: Pro is around $20/mo; Max 5x is around $100/mo for roughly five times the Pro session usage; Max 20x is around $200/mo for roughly twenty times Pro per session; Free is variable and limited. Check official Claude pages before purchase, procurement or sensitive uploads.
| Plan | Best for | How limits usually feel | Practical note |
|---|---|---|---|
| Free | Trying Claude, occasional questions, writing drafts, basic analysis. | Lower and variable. Free access can stop sooner during heavy usage or long conversations. | Good for testing Claude, not reliable for heavy daily document work. |
| Pro | Regular individual use, writing, analysis, projects, research and Claude Code access. | Anthropic says Pro has usage limits; usage depends on message length, attachments, conversation length, model and feature used. | Use shorter chats, projects and fewer repeated uploads to stretch usage. |
| Max | Heavy individual users who regularly hit Pro limits. | Anthropic positions Max as more usage than Pro, with higher output limits and priority access at busy times. | Useful when Claude is central to daily work and Pro is too tight. |
| Team | Small and mid-size teams using Claude together. | Team usage depends on seat type, workspace features, connectors and org settings. | Better than personal accounts when billing, admin and company data matter. |
| Enterprise | Large companies, security, procurement, access control and compliance needs. | Enterprise usage, spend controls, support and deployment are organization-specific. | Ask Anthropic or your workspace admin; do not rely on third-party quota claims. |
| API | Developers building Claude into products or internal systems. | API limits and billing are different from claude.ai chat subscriptions. | Check console limits, rate limits, token pricing and model choice separately. |
When capped
What to do when Claude says you hit a limit.
You usually have several clean options: wait, reduce context, use Projects, change model, upgrade or move the task to another legitimate AI tool.
Read the message
Claude usually tells you whether you hit a usage limit, need to wait, should switch model, or should upgrade.
Wait for reset
If the issue is a session cap, waiting for the stated reset is the cleanest path. Do not build your workflow around account tricks.
Shorten the task
Ask multiple related questions in one message, reduce unnecessary context, summarize first, and split very large work into stages.
Use Projects
Anthropic recommends Projects for related work. Cached project context can be more efficient than repeatedly uploading the same files.
Change model or feature
A lighter model or simpler mode can be enough for drafting, outlining and quick editing.
Use alternatives responsibly
If Claude is temporarily capped, try ChatGPT, Gemini, Perplexity, Copilot, Poe or an all-in-one tool like MultipleChat. Running Claude alongside other models means one app’s cap does not have to block you for the rest of the day.
Real situations
The right move depends on what you were doing.
A long writing job, an agent coding run, a research session across many files and a cap that lands mid-task are different problems. Treating them the same wastes time and usage. The questions worth asking are which model you really need, how much context you are carrying, and whether the work can be split so a single cap does not stop everything at once. Match the move to the situation below.
You are writing a long document
Do not regenerate the whole piece each time. Ask Claude to summarize decisions so far, then continue section by section in a focused chat. Use a lighter model such as Sonnet for drafts and save Opus for the final polish.
You are running a coding agent
Claude Code burns the shared usage pool fast because it reads files, reasons across code, edits and runs commands. Scope each task to one issue, point it at specific files, and avoid one huge vague request that runs for a long time.
You are doing research with many files
Big PDFs and repeated uploads consume usage quickly. Put recurring materials in a Project so context is reused instead of re-uploaded, and ask for an outline before deep section-by-section analysis.
You hit the cap mid-task
You usually do not lose your work. Ask Claude to write out all decisions, open questions and next steps, copy that summary, then resume in a new chat after reset or on another model so nothing is lost.
You are close to a weekly limit
The five-hour limit can clear while the weekly limit is still tight. Check Settings → Usage to see which layer is the bottleneck, and shift lighter work to Sonnet or Haiku to protect Opus weekly usage.
You need to keep going right now
If the task is urgent and the cap is firm, move the work to another legitimate model rather than waiting. An all-in-one tool like MultipleChat lets you continue alongside other models without an account trick.
Claude Code
Claude Code limits are not just chat messages.
Claude Code can work inside terminals and repositories. Coding-agent work can consume usage differently because it reads files, reasons across code, edits, runs commands and may use many tool calls. It shares the same usage pool as Claude chat and Cowork, so a long agent run draws down the same limits as your conversations. As of May 6, 2026, Anthropic doubled the five-hour limits and removed the earlier peak-hour reductions, but you should still verify your current limits in Settings → Usage.
Claude chat
Best for prose, analysis, brainstorming, summaries, document review and general questions.
Claude Code
Best for repository work, terminal tasks, code edits, debugging and implementation workflows.
Claude Cowork
Useful when you want a collaborative coding workflow and richer project reasoning, where available.
Deep dives
Claude limit guides by plan and task.
Dedicated pages for each plan and the most-searched limit questions.
Claude Pro limits
How Pro caps feel, the 5-hour and weekly layers, how to stretch usage.
Claude Max limits
Max 5x vs 20x, the two weekly limits, when Max beats Pro.
Claude Code limits
Why agent work burns the shared usage pool faster than chat.
Claude Cowork limits
Cowork shares your plan's usage with chat and Claude Code.
Claude Free limits
What the free plan includes and why it stops sooner.
Claude message limit
How many messages you actually get per plan.
Claude weekly limit
The weekly cap on paid plans, on top of the 5-hour limit.
Claude limit reset
When the 5-hour and weekly limits reset.
Claude API rate limits
RPM/TPM, usage tiers and how API limits differ from chat.
Claude Team & Enterprise
Seats, admin controls and data terms for teams.
Claude vs ChatGPT limits
How the two handle caps and pricing, side by side.
How to avoid Claude limits
A legitimate playbook to get more done without account tricks.
Myths
Bad advice about Claude limits creates account risk.
The safest playbook is boring but useful: read the in-product message, use official plans, reduce heavy tasks and wait for reset. Tricks that promise “unlimited” Claude usually create terms, billing or account problems instead, and they rarely raise your real usage. Here are the claims worth ignoring, and what to do instead.
“Make a second account for more usage.”
Building a workflow around extra accounts to dodge caps can violate Anthropic’s terms and trigger account review. The legitimate path is to wait for reset, upgrade or reduce the task.
“A VPN unlocks higher limits.”
Do not rely on location tricks. Region availability, billing and account rules can apply, and you can create new problems without raising your real usage.
“There is an unlimited-Claude hack.”
No plan is truly unlimited, and tools or scripts that claim to remove caps usually break terms or expose your account. Anthropic states usage limits apply on every plan.
“Max removes all limits.”
Max gives far more usage than Pro, but fair-use and weekly limits still apply. It raises the ceiling; it does not delete it.
“Sending the prompt again will fix it.”
Often the opposite. Repeating a large prompt burns more usage. Summarize, narrow the task and ask for one specific output instead.
“The legitimate options are worse.”
Waiting for reset, choosing a lighter model, using Projects, upgrading or running another model in parallel are all safe and keep your account in good standing.
Prompt fixes
Use leaner prompts before you burn usage.
When you are close to a cap, the goal is not to send more. It is to preserve context, narrow the work and make each model call count.
Batch related questions
Ask several connected questions in one message instead of many separate turns. One focused prompt usually costs less usage than a long back-and-forth.
Trim the context
Remove pasted text and instructions Claude no longer needs. Long, stale context rides along on every turn and quietly drains your session.
Summarize, then continue
Ask Claude to summarize decisions, open questions and next steps, then start a fresh focused chat from that summary instead of dragging one huge thread forever.
Avoid re-uploading files
Do not attach the same PDF or dataset again and again. Re-extraction is expensive; reference what is already in the conversation or a Project.
Pick Sonnet or Haiku for drafts
Lighter models are enough for outlines, first drafts and quick edits. Reserve Opus for the hard reasoning where it actually changes the result.
Use Projects for repeated work
Anthropic recommends Projects for related materials. Cached project context is more efficient than pasting or uploading the same files into every new chat.
Team & Enterprise
Company Claude limits are about more than messages.
For teams the important questions are rarely a public message number. They are seats, workspace usage, admin control, data policy and procurement, plus when people should stop relying on personal accounts at all. A single heavy user hitting a personal cap is an individual problem; a whole team sharing materials, billing and compliance obligations is a workspace problem, and it is usually solved by the plan structure rather than by chasing more messages.
Seats and shared usage
Team plans bill per seat, and usage depends on seat type, workspace features and connectors. A team workspace makes billing, ownership and shared limits clearer than a pile of personal Pro accounts.
Admin controls and data policy
Admins can govern connectors, allowed features, retention and sharing, and Anthropic offers different data-handling terms for business plans. The account cap may not be the only limit users feel.
Enterprise and procurement
Enterprise decisions center on security review, access management, support, spend controls and contract terms. Limits there are organization-specific, so confirm with Anthropic or your account team rather than a third-party page.
When to move off personal accounts
Once company data, compliance or shared billing matter, move to Team or Enterprise. Ask an admin before uploading private or regulated files to a personal Claude account.
Restrictions
Is Claude restricted? Policy vs usage limits.
Claude has two different kinds of limits: usage caps that count how much you can do, and policy or safety rules about what it will do. They are easy to confuse because both can stop a conversation, but the fix is different for each. A usage cap clears with time, a lighter model or an upgrade; a policy decision does not, no matter how long you wait or how much you pay.
A refusal is not always a usage cap
If Claude declines a request, the reason may be policy, safety or missing context rather than a quota. A refusal is a different situation from “you have reached your limit,” and waiting for reset will not change it.
Paying more does not remove rules
Pro and Max raise how much you can use, but they do not unlock restricted content or override safety policy. Plan level changes usage access, not Anthropic’s rules.
Sensitive and 18+ requests depend on policy
Adult or sensitive topics are handled according to Anthropic policy, context, legality, safety, minors and consent. Do not treat any third-party page as permission; check official policy for what is allowed.
Read the exact message first
Before assuming a daily cap, read what Claude actually said. A wait time, a model switch, a policy note and a capacity message all mean different things and point to different fixes.
Sources
Where to verify current Claude limits.
Current limits can change. Use official pages first, then compare alternatives only after you understand your actual bottleneck.
Claude app
https://claude.aiOpen Claude and check the account message shown inside the product.
Claude pricing
https://claude.com/pricingOfficial plan names, current plan details and usage wording.
Claude Pro usage
https://support.claude.com/en/articles/8324991-about-claude-s-pro-plan-usageAnthropic explains that Pro has usage limits, resets and variables.
Claude Code
https://claude.com/product/claude-codeClaude Code is developer-agent work, not only ordinary chat.
Anthropic usage policy
https://www.anthropic.com/legal/aupPolicy restrictions are separate from usage caps.
FAQ
Claude limit questions answered.
Short answers for people searching Claude daily limit, Claude Pro limit, Claude Max limit, Claude Code limits and free alternatives.
Does Claude have usage limits?
Yes. Claude plans can have usage limits. Anthropic says message count varies based on message length, files, conversation length, model and feature used.
How many Claude messages do I get?
There is no single stable number for every user. Anthropic gives examples for relatively short conversations, but actual usage varies by model, feature, attachments and capacity.
When do Claude limits reset?
Anthropic says Pro session-based usage limits reset every five hours. Other caps, plan rules or feature limits can work differently.
Why did Claude limit me faster than expected?
Long files, long chats, repeated uploads, expensive models, Research, code tasks and high traffic can all consume usage faster.
Is Claude Pro unlimited?
No. Pro gives more usage than free access, but Anthropic explicitly says usage limits apply.
Is Claude Max unlimited?
No. Max is positioned for more usage than Pro, but limits and fair-use rules still apply.
Are Claude Code limits the same as normal Claude chat?
No. Claude Code is developer-agent work. Anthropic says Pro users can use Claude Code and that usage limits apply across Claude and Claude Code.
Can I bypass Claude limits with another account?
Do not build around account tricks. That can create account and terms risk. Use official plans, wait for reset, reduce the task or use legitimate alternatives.
Does Claude allow 18+ content?
Policy is separate from usage limits. Sensitive or adult requests depend on Anthropic policy, context, legality, safety, minors and consent. Check official policy.
Can Claude read unlimited PDFs?
No. Files, context, extraction quality, conversation length and analysis effort all matter. Very long documents can use more of your limit.
How do I make Claude usage last longer?
Use concise prompts, combine related questions, avoid re-uploading files, use Projects for repeated materials and start focused new chats when old ones get huge.
What should I use when Claude is capped?
Wait, switch to a lighter model, upgrade, or use legitimate alternatives such as ChatGPT, Gemini, Perplexity, Copilot, Poe or MultipleChat.
Is MultipleChat a Claude replacement?
It is better described as an all-in-one AI workspace. You can compare Claude-style answers with other models and keep working when one model is not the best fit.
What is the difference between the weekly and five-hour reset?
There are two layers. The five-hour session limit resets about five hours after your first message. Paid plans also carry weekly limits that reset on their own schedule, with a separate reset for Opus. Check Settings → Usage to see which one is the bottleneck.
Does Claude Code share my chat limit?
Yes. Claude Code draws on the same usage pool as Claude chat and Cowork. A long agent run that reads files, edits code and runs commands can use that shared usage faster than ordinary conversation.
Why did Pro stop me so fast?
Long files, long threads, repeated uploads, heavier models like Opus, Research and code tasks all consume usage faster. A short session limit plus a tight weekly limit can also combine. Verify both in Settings → Usage.
Do attachments cost more of my usage?
They can. A short message and a large PDF, dataset or code archive do not place the same load on the system. Big or repeated uploads use more of your limit than plain text.
Is Max 20x worth it?
It depends on your workflow. Max 20x is around $200/mo for roughly twenty times the Pro session usage; Max 5x is around $100/mo for about five times. If you regularly hit Max 5x limits during real daily work, the higher tier can be worth it. Verify current pricing officially.
Can I buy extra usage when I run out?
Options and add-ons can change over time, so check Settings → Usage and the official pricing page for what is offered on your plan. Upgrading a tier, waiting for reset or using a lighter model are the reliable paths.
What is the API rate limit versus chat?
They are separate products. The Anthropic API has its own rate limits, token pricing and billing in the console, which are different from claude.ai subscription usage. Check console limits independently from your chat plan.
Does a refusal mean I am capped?
Not necessarily. A refusal can be a policy, safety or missing-context decision rather than a usage cap. Waiting for reset will not change a policy refusal; read the exact message to tell them apart.
Why did Claude switch me to a lighter model?
When usage is tight or demand is high, you may be offered or moved to a lighter model such as Sonnet or Haiku. That is normal limit behavior and is often enough for drafting and quick edits.
Should I pay just because I hit one limit?
Not always. If it happened once, wait or reduce the task. If a cap blocks normal daily work, compare Pro, Max 5x and Max 20x by your actual workflow, not by one frustrating moment.
Who has the final word on Claude limits?
Claude.ai, Claude pricing, Anthropic support articles, Anthropic policies and your workspace admin have the final word. This page can be wrong or outdated.