At a Glance · AI-generated

Anthropic released Claude Opus 5 to developers on July 24, maintaining the same pricing as its predecessor while introducing enhanced capabilities in agentic coding and computer use. The model demonstrated improved performance on cybersecurity tasks and achieved top benchmark scores against competitors including Fable 5 and GPT-5.6 Sol. According to reports, Opus 5 addressed a critical security vulnerability affecting AI agents by mitigating browser-based prompt injection attacks.