Make Claude Feel Limitless

Six practical ways to get more done before you hit your usage limit

Hey smarties 👋

Claude can feel generous until you hit your limit halfway through a project. Often, it isn’t the project itself that drains your allowance. It’s the settings you choose and the habits you barely notice. This week, we’re looking at what uses up your Claude limit and six ways to make it last longer.

Sponsored

Stop betting on one AI answer

Sponsored

According to Artificial Analysis' AA-Omniscience benchmark, the best AI model is right only 67% of the time. And it sounds confident even when it's wrong.

Cuey is a free Chrome extension that gives you a second opinion, so you don't get burned by the first one.

  • Cuey lives inside ChatGPT, Claude, Gemini, Grok, and more.

  • One prompt, three answers, side by side. No tab-switching, no copy-paste.

  • See where the models disagree before a hallucination costs you.

  • Carry your context across 30+ models, so you never start from scratch.

The Surprise

Is Claude Actually the Better Deal?

Claude looks expensive if you judge it one task at a time. In a cost comparison, a demanding task cost $0.21 with GPT-6.1 Sol and $1.34 with Opus 5.5.

You might expect Claude’s $100 plan to run out first. But SemiAnalysis tested the two subscriptions and found the opposite. It priced the work each plan allowed at API rates:

  • ChatGPT’s $100 plan: $1,055 worth of work

  • Claude’s $100 plan: $5,725 worth of work

How can the pricier model give you more work? Because a subscription doesn’t charge you for each task. You pay a fixed monthly price, and the company decides how much usage to include. In this test, Anthropic gave its $100 subscribers a much larger allowance.

For someone using Claude every day, the question becomes how to make that allowance last. That’s what this guide is about.

UNDER THE HOOD

What Uses Up Your Claude Limit?

Claude doesn’t only count the answer you see. It also has to process what you send before it can reply. That’s why a quick question at the end of a long chat can use more of your allowance than you expect.

Think of it in three parts:

  • Reading: Your message, the earlier conversation, and any files you’ve shared.

  • Thinking: Working through the task. A higher effort setting generally uses more.

  • Writing: Producing the answer, document, or other result you asked for.

Claude can reuse some recently processed material through a cache, which helps. But the simplest way to make your allowance last is to give it less unnecessary material to read, use the right effort setting, and ask for the output you actually need.

THE PLAYBOOK

6 Ways to Make Claude Feel Limitless

01 / MODEL & EFFORT

Choose Your Settings Before You Start

Use Medium Effort. Almost Always:

Effort controls how much reasoning Claude puts into a task. More effort can improve the result, but it also uses more tokens.

Look at the chart: the score rises as you move up, but so does the cost. My rule:

  • Low for grammar, reformatting, and translations.

  • Medium for almost everything else.

  • High for a one-shot prompt on a particularly ambitious task.

  • Never Extra. Never Max as your everyday setting.

Use Opus 5.5 as Your Default:

On Claude’s $100 plan, the estimated monthly value was $5,725 with Opus 5.5, compared with $1,273 with Fable 5.1 before Fable reached its separate limit. Sonnet 5.5 was close to Opus at $5,919.

For most chats, start with Opus. Need a quick answer? Lower the effort. Save Fable for work where you actually need it.

Don’t Switch Models Mid-Chat:

Each model has its own cache. Switch from Opus to Sonnet halfway through a chat, and Sonnet has to read the conversation again. That uses more of your limit.

Pick Opus 5.5 when you start. Stick with it until the task is done.

02 / CHAT HABITS

Keep Your Chats Short

Don’t Return to a Long Chat Hours Later:

After a long break, Claude may have to read the old conversation again. Before you leave, ask it to summarize what you decided, what’s unfinished, and what to do next. Paste that summary into a new chat when you return.

One Topic Per Chat:

If you’re done with a proposal and want help with a presentation, start a new chat. Claude doesn’t need the entire proposal conversation to work on something else.

Put Related Requests in One Message:

Need a summary, key points, and a draft email? Ask for all three together. Three separate follow-ups make Claude work through the growing chat three times.

03 / KEEP IT FOCUSED

Don’t Make Claude Do Extra Work

Keep Quick Questions in Chat:

If you only need a short answer, say so: “Answer here in under 150 words. No web search or files.” Save the extra tools for questions that actually need them.

Plan First. Make the File Last:

Don’t start a presentation by asking Claude to build the whole deck. Work out the outline and wording in chat first. Once you’re happy with them, ask for the file.

Change One Part, Not Everything:

If one paragraph in a document is wrong, select it and ask Claude to fix that part. In chat, you can say: “Rewrite only the second paragraph. Keep everything else.”

04 / WHAT CLAUDE READS

Give Claude Less to Carry

Load Apps Only When Needed:

If you have lots of connected apps, set Tool access to On demand. Claude will look for the app it needs instead of loading them all into the chat.

Crop Screenshots. Share the Relevant Text:

Asking about one error? Crop the screenshot to that error. If you only need help with a few paragraphs from a PDF, share those paragraphs. Keep the full file when its charts or layout matter.

Put Reused Files in a Project:

Use the same brand guide or research notes across several chats? Add them to a Project once instead of uploading them each time. Project content can be cached when you reuse it.

05 / INSTRUCTIONS

Tell Claude What Matters Up Front

Keep Your Standing Instructions Short:

Your Instructions for Claude apply across chats. Keep the preferences you always want there. Put the details of today’s task in today’s message.

Stop It Reworking Finished Answers:

For everyday chats, add this instruction: “On follow-ups, focus on my new request. Revisit an earlier answer only if I ask or you notice a mistake.” That keeps a small follow-up from turning into another long answer.

Let Claude Ask Before It Starts:

If an important task is still vague, don’t spend ages guessing what to put in the prompt. Give Claude the goal, then say: “Ask me what you need to know before you begin.”

05 / INSTRUCTIONS

Tell Claude What Matters Up Front

Keep Your Standing Instructions Short:

Your Instructions for Claude apply across chats. Keep the preferences you always want there. Put the details of today’s task in today’s message.

Stop It Reworking Finished Answers:

For everyday chats, add this instruction: “On follow-ups, focus on my new request. Revisit an earlier answer only if I ask or you notice a mistake.” That keeps a small follow-up from turning into another long answer.

Let Claude Ask Before It Starts:

If an important task is still vague, don’t spend ages guessing what to put in the prompt. Give Claude the goal, then say: “Ask me what you need to know before you begin.”

06 / USAGE CONTROL

Don’t Burn Your Limit Unnoticed

Check Usage Before a Big Task:

Claude has a five-hour limit and a weekly limit. Before starting a long job, check Settings → Usage. If the five-hour limit is nearly gone, save that job for after it resets.

Pause Scheduled Tasks You Don’t Read:

An hourly task can run 24 times a day, even when your laptop is closed. Open Scheduled, look at what’s still running, and pause anything whose results you no longer use.

Save a Free Reset for When You Need It:

If Settings → Usage shows a free reset, don’t use it for a routine question. Keep it for a deadline when you’ve hit a limit and check whether the offer has an expiry date.

07 / READER POLL

Last Edition’s Poll Results

Thank you to everyone who voted. Here’s how you rated the last MonDive.

Amazing, very clear
66.7%

Good, but a few parts were tricky
16.7%

Not impressive, it still felt overwhelming
16.7%

A NOTE FROM A READER

Selected: Amazing, very clear

“Love it”

Our reply: Thank you! That’s lovely to hear. We appreciate you reading MonDive and taking a moment to share your feedback.

FROM SMARTER WITH AI

Clear your AI backlog. Turn saved ideas into useful work. Practical guides, walkthroughs and adaptable prompts in the AI Library. $29 once, no recurring subscription.

Practical guides, walkthroughs and adaptable prompts to help you put AI to work. $29 once. No recurring subscription.

Explore the AI Library →

We’d love to hear from you!

How did you feel about today's MonDive? Your feedback helps us improve and deliver the best possible content.

Login or Subscribe to participate

Know someone who may be interested?

Thanks for reading. See you next Monday.

Reply

Avatar

or to participate