Breaking

Tom Dörr Highlights 30-Word Claude Prompt as Concision Tradeoff Resurfaces

September 24, 2026 at 00:47 EDT

Foundation Models
Software Dev & Coding
AI Agents

Tom Dörr's under-30-word instruction for Claude has put a fresh spotlight on a practical way users are trying to cut AI chat overhead. The tactic is not a new Anthropic release, but it lands in a live debate over how far concise prompting can go before it starts to damage answer quality.

Sep 24, 2026 · Claude

The 30-word Claude shortcut

A strict reply cap can trim chat overhead, but prior Claude Code tests show that shorter is not always smarter.

<30

words per reply in the user rule

100

word Claude Code cap later reverted

drop in one broader eval

output-token price ratio on listed current models

Word-count benchmark

A terse-rule test reported total output falling from 465 words to 170 words.

465 words

Baseline

170 words

Optimized

Takeaway: the reported word-count drop was large, but still directional and workload-dependent.

How the workflow changes

Set brevity instruction

→

Ask narrow question

→

Get terse answer

→

Expand when judgment matters

Where it helps

Fast factual lookups, code-review nits and repeated agent loops can benefit from less filler and lower generated text.

Where it bites

Tradeoffs, debugging narratives and recommendations can lose the rationale that makes the answer useful.

The editorial bottom line

Make brevity the default for simple exchanges, but leave the model room when the reasoning is the product.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.