Claude hit the maximum length for this conversation

Claude told you the chat has reached its maximum length and won't continue. It's a hard limit, not a fault — and your work is still there. Here's what it means and how to pick up in a fresh chat without re-explaining everything.

What the limit actually means

Each time you send a message, Claude re-reads the entire conversation — every message from you, every reply, and anything you pasted or uploaded. All of it has to fit inside a fixed budget called the context window. Once the thread grows past that budget, there's no room to add a new message alongside the full history, so Claude stops and tells you the conversation has reached its maximum length.

It tends to hit at the worst moment: a long document you've been drafting, a deep back-and-forth, a project built up over several sittings. Nothing is broken and nothing is lost — the thread simply got too big to continue in place. The way forward is a new chat.

The wrong ways to deal with it

Two moves feel natural, and both cost you time:

• Retyping everything from memory. Slow, and you inevitably drop the details that mattered most — the constraints you kept repeating, the direction you already ruled out.
• Pasting the whole old chat into a new one. This just recreates the wall you hit. A giant block of history fills the fresh budget straight away, and you're back at the limit within a handful of messages.

The right way: write a handoff

Rather than dumping the full history, give the new chat a tight briefing — only the parts that still matter. Four moves:

1. State the goal in one line. What are you producing? Be concrete: "a research memo on X", "a rewrite of the intro", "a working version of this function".
2. List the locked decisions. What you already agreed and don't want reopened — the structure, the tone, the approach you set aside.
3. Write down the constraints you kept repeating. If you reminded Claude of the same rule over and over ("keep it concise", "British spelling", "no headings"), those are the first things a fresh chat won't know. Spell them out.
4. Capture where you left off. Paste the latest draft, or the exact question you were mid-answer on.

Then open a fresh chat and paste it as the first message. A template you can copy:

Pick up exactly where we left off. Everything below is settled — don't re-ask it.

GOAL: [what we're producing, in one line]

LOCKED DECISIONS:
- [decision 1]
- [decision 2]

CONSTRAINTS (keep these every time):
- [constraint 1]
- [constraint 2]

WHERE WE LEFT OFF:
[paste the latest draft, or the exact question we were on]

Continue from here.

A note on sharing your Claude chat

If you want a tool to read the conversation from a share link, Claude requires you to set the share to "Anyone with the link" first — that's the only way anything outside your account can open it. That's a real trade-off worth knowing: while it's shared, anyone who has the URL can view that chat. The safe habit is to share, use the link, then delete the share once you're done. If you'd rather not share at all, you can always copy the conversation text and paste that instead.

The catch with doing it by hand

Writing the briefing yourself works, but it's slow — and you tend to forget the constraints you were most tired of repeating, which are exactly the ones the new chat needs. That's the job Uncook automates.

Related: ChatGPT: "conversation is too long" · Why AI chats get worse over time