OpenAI and Anthropic are no longer racing on capability alone, they are racing on price, and Anthropic's new model is the latest proof.
The AI company released Claude Opus 4.8 Thursday, and it is faster, better at coding, and priced exactly the same as the previous version at the standard tier.
What Is Actually New
Anthropic says Opus 4.8 outperforms competitors on benchmarks for agentic coding, reasoning, financial analysis, and "knowledge work," which is the kind of long, multi-step task real companies are trying to hand off to AI.
There is also a new fast mode that runs 2.5x quicker than before at about one-third the cost of earlier fast modes.
And a new "effort" control panel lets users decide how hard Claude thinks about a problem, with lower effort burning fewer tokens while higher effort thinks longer.
Anthropic also rolled out a "dynamic workflow" feature that lets one Claude run multiple sub-agents at once, which is the kind of feature enterprise customers have been pushing for.
Same sticker price, faster output, and more control over the bill is a quiet win for any company writing a check to Anthropic.
If you want a daily handle on which AI moves actually matter for the broader market, Market Briefs breaks it down every morning, with a free investing masterclass included.
The Race Behind The Race
Pricing matters because AI bills have moved from a "nice to have" to a real line item, and customers want to spend less, not more.
Anthropic is bundling features that read as enterprise productivity plays, with the effort dial letting developers cap costs without giving up quality on the things that matter.
Competitors are not standing still, and OpenAI has been cutting prices on its lower-tier models for months, with the result that capability gaps are getting smaller while pricing gaps do more of the work.
Anthropic's response is to bundle features that justify the price tag without raising it. Same input cost, more output value, which is the math customers have been asking for.
What To Watch
Opus 4.8 still trails Mythos, Anthropic's most powerful model, which was released initially to a handful of partners because of security concerns.
Anthropic says a Mythos-class model will be available to all customers "in the coming weeks."
When that lands, the comparison gets sharper. Until then, the headline is the price tag.
The IPO market is also watching, because Anthropic is reportedly in line for one of the most anticipated AI listings, and pricing strategy is part of the story investors care about.
For more on the AI race and what it actually means for your portfolio, sign up for Market Briefs. You also get a 45-minute investing course as a bonus when you join.
