Here goes the video

Nah, this is not correct. No AI tool/model work this way. With this sort of prompts even if you use different chat bubbles you won’t hit limit that easily. Ask it do a serious task even with this style and you’d hit limit in just 4 to 5 requests (on free plan)

Lmao you just don’t use Claude for complex tasks… that usage limit is inevitable

@TheRuqayyah Nawa oo. 😭 btw this will only work for research, and not coding or even design-related tasks.

@Toluwan29110384 Wrong o, editing still costs tokens both input and output. Then summarizing to go to a new chat is a waste of token too 😭

@TheRuqayyah Claud when I tried the all in one prompt for design tasks 😭

@TheRuqayyah 😅😅😅😅

@TheRuqayyah 😅

@Crownzdesigns

Not entirely true. These type of tasks are not heavy. One will still eventually run out of tokens. One thing users forget is the starting point of prompting the AI to be efficient in its output; you can mention to avoid exhausting tokens easily, and also providing laid out writing guidelines. What users often forget: Input tokens count too. If you paste a 10-page doc every turn, you'll exhaust tokens no matter how short my reply is. The starting point matters on both sides. Quick tips to stretch tokens: Be specific: "Explain RAG in 3 bullets" vs "Tell me about RAG" Set limits: "In 50 words or less" Iterate: Ask for the outline first, then expand only what you need Reference, don't repeat: Say "use the code from before" instead of pasting it again.

@DabereNnamani 🍀

@DabereNnamani

This looks helpful. Thanks Jennie

Thanks, that was helpful @jenidesignns

@Oyin_Laoye

Totally wrong instructions. There are cached tokens, it’s better you stay on your chat session while Claude cache previous messages than copying summary to a new chat.

You will still get the limit because it has current session limit even for the pro. Try using it for coding...

Everything we watched in movies All playing right

Thanks for sharing

The thing go hallucinate

loading...

Omo eh

Hi @jenidesignns editing a message still costs tokens 👀. Claude discards everything after the edit and restarts from that point, resending the full conversation history + your new content. It can actually cost more than just sending a new message.

Thank you for sharing

This will only work for research, not vibe coding

How did you guys open Claude with Nigeria number

Select your country

I did select my country but it’s not sending me an otp code

1. Edit previous prompt if you’re unsatisfied, don’t use a follow-up prompt. 2. If conversation is getting too long, ask for summary and start a new chat with the summary. 3. Ask all your questions in one prompt. Don’t separate them into multiple prompts. 4. Use cheaper models

Oh wow insightful. Thanks a bunch

Partially correct, but for web or mobile. For engineers and executors, this won't work

This doesn't solve a problem where you only ask one question for the first time that day, and Claude says you've resahed your limit (without answering the question).

Wont Work

Thank you for sharing

also, screenshots and files burn your tokens faster

Interestingly, I have been adopting this trick subconsciously without knowing that, I was saving my token. Thanks for this refresher video.