The Frontier Gets Cheaper: Opus 5.5 and GPT-6 Variants Compete on Price, Not Just Capability

Anthropic's Claude Opus 5.5 reportedly costs 40% less than its predecessor while OpenAI's GPT-6 line pushes the same lever, marking a shift in how the labs are positioning their newest frontier models.

The competitive story at the frontier is quietly changing. In a rundown of recent releases, @zachdavis noted that Claude Opus 5.5 "costs 40% less" than its predecessor, and framed the current OpenAI GPT-6 lineup — the Sol and Luna variants — as "advancing the frontier on cost efficiency." That is a meaningful rhetorical shift. For much of the last two years, frontier releases were sold on raw capability. Now the pitch increasingly leads with price-performance.

Unlock the full briefing

Get every story in today's briefing, the full archive, and the daily AI intelligence brief.

All stories today

Full archive

Daily brief

Cancel anytime. Payments powered by Stripe.