Anthropic released Claude Opus 5 on July 24, an updated hybrid reasoning model that delivers stronger coding, knowledge, and agent performance at the same price as its predecessor. The model nearly matches the company's top-tier Claude Fable 5 on several benchmarks while costing half as much, making frontier-level AI more accessible for developers and IT professionals.
Benchmark performance
On Frontier-Bench v0.1, Claude Opus 5 more than doubled Opus 4.8's score and topped all other models tested. On CursorBench 3.2, a coding evaluation, it came within 0.5% of Fable 5's peak score at 50% lower cost per task. Across high, xhigh, and max effort settings, Opus 5 outperformed all competitors at equivalent costs.
Anthropic described the model as "a thoughtful and proactive model" that comes close to the "frontier intelligence" of Claude Fable 5 at half the price. Opus 5 is now the default on Claude Max and the strongest model available on Claude Pro.
Pricing and developer access
Claude Opus 5 costs $5 per million input tokens and $25 per million output tokens, unchanged from Opus 4.8. Developers can call it through the Claude API. Two accompanying beta features-mid-conversation tool changes on the Claude Platform and automatic fallbacks on the API-aim to improve reliability and flexibility for technical teams.
Why this matters for IT and Development
For developers and IT engineers, Claude Opus 5 offers a notable upgrade in coding and agentic tasks without increasing cost. The near-parity with the top model on CursorBench means teams can integrate a high-performing reasoning engine into automated workflows, code generation pipelines, and internal tools at a lower API bill. The new fallback and tool-change features further reduce downtime and manual intervention, helping teams build more resilient AI-driven applications.
Your membership also unlocks:
