Anthropic unveils latest AI model, aiming to extend its lead in coding intelligence (www.businessinsider.com)

Anthropic unveiled Claude Sonnet 4.5, positioning it as the company’s new flagship for AI coding and applied intelligence just four months after Sonnet 4. The model claims state-of-the-art results on SWE-Bench Verified and boosts code reliability, refactoring judgment, and production-readiness while adding cybersecurity capabilities (faster vulnerability detection and remediation). A standout technical claim is a dramatic increase in autonomy—Sonnet 4.5 can operate independently for up to 30 hours, more than quadrupling the endurance of Anthropic’s prior Opus 4—making it suited for long-running, end-to-end engineering workflows. Anthropic says Claude Code is already generating over $500M in run-rate revenue with usage growing 10x in three months, underscoring that coding is a core commercial driver. Alongside the model, Anthropic launched developer-focused infrastructure and an SDK for Claude Agents to build customized, context-aware AI agents: virtual machines, memory and context management, a native VS Code extension, enhanced terminal workflows, and checkpoints to roll back code. Those tools aim to accelerate production-ready app creation and wider adoption of autonomous developer agents, intensifying competition with Google’s Gemini, OpenAI’s GPT-5, and xAI’s Grok 4. For the AI/ML community, Sonnet 4.5 signals faster iteration on capability and tooling that could materially shift software engineering productivity, security automation, and how teams integrate persistent, context-rich agents into production.