Anthropic launches Claude Opus 4.5 flagship AI model
Anthropic launches Claude Opus 4.5 flagship AI model
Anthropic introduced Claude Opus 4.5, a new flagship model aimed at software development, agent architectures, and complex technical tasks.
Availability and platforms
The company made the model accessible in its applications, via API, and on three major cloud platforms to support diverse deployment scenarios.
Technical improvements
Opus 4.5 demonstrates increased stability, enhanced resistance to prompt injections, and an ability to solve tasks that earlier releases could not complete reliably.
Anthropic reports record benchmark scores across several evaluation suites and highlights improved behaviour in multi-system and agent-based workflows.
Pricing and positioning
Pricing is set at $5/$25 per million tokens, which Anthropic cites as positioning Opus as a more accessible tier-one model for developers and enterprises.
Implications for developers
The vendor expects the release to address common engineering requirements for code generation and automated agents while reducing operational risks from adversarial prompting.
Access through APIs and major clouds aims to simplify integration into existing toolchains and to support scalable workloads across organizations.
Related posts
