Anthropic released Claude Opus 5 on Tuesday, a new flagship AI model that the company says delivers stronger coding and knowledge-work performance than its predecessor — and matches a top competitor at roughly half the cost.
What Opus 5 brings to the table
The model outperforms Opus 4.8 on tasks that involve writing code, analyzing documents, and synthesizing information, according to Anthropic. It's designed for professionals who need reliable, high-level reasoning rather than just chat. The company positioned it as a direct alternative to Fable 5, another leading model on the market. Internal benchmarks show Claude Opus 5's capability landing near that of Fable 5, though Anthropic didn't claim outright superiority.
Pricing that undercuts the competition
Anthropic kept the price for Opus 5 the same as Opus 4.8. That means users pay half of what Fable 5 costs for comparable performance. The move signals a clear strategy: compete on both quality and cost. For businesses burning through API calls on complex tasks, the savings add up fast.
The pricing isn't a temporary promotion. Anthropic set it as the standard rate for the new model, making it a permanent option in its lineup.
Why the coding focus matters
Anthropic's emphasis on coding and knowledge work reflects where the AI market is heading. Companies want models that can actually generate production-ready code or summarize dense research, not just answer trivia. Claude Opus 5's improvements in those areas target developers, analysts, and researchers — the people most likely to pay for premium access.
The model also handles longer contexts and more complex instructions than earlier versions, though Anthropic didn't release exact token limits in the launch materials.
The competitive landscape
Fable 5 has been the benchmark for high-end AI performance since its release last year. By pricing Opus 5 at half that, Anthropic is betting that many customers will switch if the quality gap is small enough. Early testers reported that Opus 5 handles debugging and multi-step logic tasks with fewer errors than Opus 4.8, but independent third-party evaluations are still pending.
Anthropic didn't say whether Opus 5 will eventually power a consumer chatbot or remain strictly an API product. For now, developers can access it through the Anthropic API and the company's console.
