Claude Opus 5.5 Cuts Em Dash Use by 95% as Responses Grow Longer
Sign in✓ Signed in
Claude Opus 5.5 is drawing attention for a measurable style shift: independent analysis found 95% fewer em dashes than Opus 5, even as average answers grew 6% longer. The finding lands just days after Anthropic released its new flagship Opus model for long-running coding, agentic work and knowledge tasks.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...