Anthropic Launches Claude Opus 5, Undercuts Rival on Price

What happened

Anthropic has released Claude Opus 5, positioning the new model as a "thoughtful and proactive" system that comes close to matching a rival model referred to as Fable 5, while costing about half as much to use 12. The announcement, dated Friday, July 24, was reported from San Francisco, where Anthropic framed the launch as part of a broader push to make advanced AI capability more affordable 2.

Both accounts describe the same core pitch: Opus 5 is not necessarily the most powerful model on the market, but it is positioned as a value play, delivering performance that approaches a costlier competitor at a fraction of the price 12. That framing matters because it signals Anthropic is competing as much on economics as on raw capability, a strategy that has become increasingly common as AI labs try to differentiate themselves beyond simple benchmark leadership.

Why it matters

The release lands squarely in the middle of what one outlet explicitly calls an intensifying AI price war 2. Coverage of Opus 5 suggests Anthropic is trying to reset expectations around what buyers should pay for near-frontier performance, rather than simply chasing the top of capability leaderboards. For businesses and developers building on Claude models, a cheaper option that claims to approach the performance of a more expensive rival could reshape cost calculations for large-scale deployments, especially for uses where marginal gains in capability matter less than throughput and budget.

The comparison point in both reports is a model called Fable 5, described as Opus 5's more powerful and more expensive counterpart 12. Neither account elaborates on whether Fable 5 belongs to Anthropic itself or to a competitor, leaving the exact competitive landscape somewhat undefined based on the reporting available. What is consistent is the framing: Opus 5 is the value-oriented alternative, priced at roughly half of Fable 5's cost while closing much of the capability gap 12.

Where the reporting agrees

The two accounts corroborate each other on the essential facts. Both confirm that Anthropic launched Opus 5, that it is described as "thoughtful and proactive," that it is pitched as nearing the capability of Fable 5, and that it is priced at approximately half of that rival's cost 12. This alignment, even though the two pieces come from different outlets, Thurrott and Yahoo's redistribution of a Reuters report, suggests the core claims trace back to a single announcement or press release from Anthropic rather than independent verification by each outlet.

Where it doesn't

The most notable difference is one of context rather than contradiction. The Yahoo/Reuters account explicitly situates the launch within a broader narrative of an intensifying price war among AI companies, a framing that Thurrott's report does not include 2. Thurrott's piece sticks closely to the product description and comparison to Fable 5 without placing it in that wider competitive narrative 1. This is not a factual disagreement, but it does mean readers get a narrower picture from one outlet and a more market-contextualized one from the other.

Neither report offers independent detail on pricing figures, benchmark scores, or specifications for Opus 5 or Fable 5, and neither identifies who makes Fable 5. Given that both articles appear to stem from the same underlying announcement, likely amplified through Reuters wire distribution, the lack of additional independent reporting means several basic questions, such as exact pricing tiers, release timing details beyond the Friday date, and what makes Opus 5 "proactive" in practice, remain unanswered by the available coverage.

The bottom line

The evidence supports treating this as a genuine product launch centered on a cost-performance argument: Opus 5 as a cheaper alternative that claims to approach a pricier rival's capabilities. The price-war framing from the Reuters-sourced report is the more useful lens for understanding why this release matters, since it connects Anthropic's move to a broader industry pattern rather than treating it as an isolated product update. Readers should note, however, that both pieces rely on Anthropic's own characterization of the model, and independent benchmark confirmation is not yet part of the public record.