Claude Opus 5 Arrives as Anthropic Challenges Fable at Half the Price

Published 31 July 2026 · Updated 19 October 2026

Anthropic just released Claude Opus 5, a new flagship model that competes directly with Fable 5 but costs a fraction of the price. Available across Claude apps, Claude Code, and the API, Opus 5 is billed as a thoughtful and proactive system that keeps Anthropic competitive in the top tier of large language models.

Anthropic’s published benchmarks put Opus 5 at the front of several categories. On agentic terminal coding, knowledge work, agentic search, and computer use, it outperforms both Fable 5 and GPT-5.6 Sol. Its ARC-AGI-3 score of 30.2 percent, reportedly verified by ARC Prize, is roughly three times the previous best. Composite rankings are more mixed: on some independent indexes it still sits behind Fable 5 and GPT-5.6.

On mathematical reasoning, the model showed perfect performance in testing against International Math Olympiad 2026 problems, scoring 42 out of 42. The previous gold threshold was 29 points.

For teams evaluating AI safety, Opus 5 also adds alignment improvements. Reviews of Anthropic’s system card suggest Opus 5 is strong at finding software vulnerabilities but is not Mythos-class at chaining them into working exploits. This boundary matters for security teams deciding which tool to trust near production code.

Price remains central to the launch. Opus 5 keeps the same rate structure as Opus 4.8: 5 dollars and 25 dollars per million tokens for input and output. That puts powerful frontier-class intelligence within reach for teams that cannot justify Fable-level pricing.

For Australian developers and businesses following the AI race, the move signals that Anthropic is no longer content with Opus sitting one tier ahead through Sonnet. Now the company is pushing down from above, using the same model family that previously competed only at the top of the market to challenge the current benchmark leaders at half the cost.

Sources

Benchmark figures are largely vendor-reported and had not been independently replicated at the time of writing.

Related Reading

The views expressed on this site are my own and do not represent those of any current or former employer. Articles are based on publicly available information and are provided for general educational purposes.

How this analysis is researched, sourced and corrected: Editorial standards and corrections. Spotted an error? Contact Philip Hall.

Subscribe

Related articles

The UK Just Narrowly Avoided a Fridge Hack. The Rest of the World Isn’t So Lucky.

A secret operation that spanned supermarkets, warehouses and stadiums In...

AI Agents Are Trading Secrets for Help. That Is a Security Architecture Problem

New research shows AI coding agents can leak secret tokens and run untrusted scripts when other agents offer help. The lesson is architectural: keep secrets out of agent runtime.

AI Agents Are Hunting Exposed Secrets, and Businesses Are Exposed

Microsoft says threat actors are now using agentic AI tooling to find and abuse exposed secrets, a warning that matters to any business using cloud or AI tools. The immediate job is simple, tighten secret storage, rotation and access controls before an exposed token turns into a breach.

CrowdStrike’s AI Agent Kill Switch Is a Sign of the Times

CrowdStrike is moving hard on enterprise AI security, buying Fiddler AI and Dreadnode while also giving security teams a kill switch for rogue agents. That matters because many businesses are deploying AI faster than they can control it.
Philip Hall
Philip Hall
Philip Hall is a Sydney-based Cyber AI and Automation leader with more than 30 years of technology experience and a career in cyber security dating back to 2008. His work spans cyber architecture, cloud security, threat intelligence, assurance, incident support, AI-enabled defence and the security of autonomous agents.