Claude says the conversation is too long. How to carry on without losing it

You are two hours into a conversation with Claude. It knows the project, the constraints, and the thing you tried on Tuesday that did not work. Then a message says the conversation has hit its maximum length, and would you kindly start a new one. The new one knows nothing. It greets you like a hotel receptionist.

First, what Anthropic's own help pages say about that limit, which has changed a good deal. Then what works on every plan and in every AI app: a handoff note, with a template to paste. I build a notes app, which turns up once, near the end, as one of four places to keep the note. Everything is linked and was checked on 12 October 2026.

📏 What the limit is

Anthropic's help centre separates two limits that people mix up. A usage limit is how much you can use Claude in a period, and it resets. A length limit belongs to one conversation and comes from the context window, which the page calls "Claude's working memory that determines how much content it can process and remember at once".

On paid plans, "the newest models support up to a 1M token context window, while others support 500K or 200K tokens." There is a table per model, which will have changed since I typed this sentence.

I could not find the message's wording on Anthropic's help pages. The version people paste into bug reports, this one from May 2025, reads: "Claude hit the maximum length for this conversation. Please start a new conversation to continue chatting with Claude." Yours may differ by a word.

🧹 What Claude now does about it unasked

The help centre now describes automatic context management: "When your conversation approaches the context window limit, Claude summarizes earlier messages to continue the conversation seamlessly." The context window article goes further: "This allows conversations to continue indefinitely in most cases."