Anthropic has launched Claude Opus 4.5, its new flagship large language model, which the company describes as its most intelligent and capable model to date.

The release is positioned as the best model yet for coding, agents and computer use, with the company claiming its agentic AI reaches peak performance after just four iterations compared to ten for rival models. On coding benchmarks, the model scores 80.9 per cent on SWE-bench Verified, according to startup benchmarks cited in launch coverage.

The launch also includes expanded context capabilities, hybrid reasoning, new integrations with Excel and Chrome, and a desktop Claude Code app for Windows and macOS. The model is available immediately on Anthropic’s platform and through major cloud providers.

Anthropic said the release underwent expanded safety testing, reflecting the company’s emphasis on responsible deployment as its models grow more capable.

The launch intensifies the race among frontier AI labs, with each new flagship aimed at winning developers, enterprises and agent builders.