Anthropic Unveils Claude Opus 4.5, Claiming Superiority in AI Benchmark Performance
Anthropic, the Amazon-backed AI firm, has launched Claude Opus 4.5, its most advanced model to date. The release follows earlier iterations, Sonnet 4.5 and Haiku 4.5, and positions Opus 4.5 as a formidable competitor against Google's Gemini 3 Pro and OpenAI's GPT-5.1. Notably, Opus 4.5 achieved an 80.9% accuracy on the SWE-Bench coding benchmark—a first for any model—earning Anthropic's bold claim of being the world's best for coding and computational tasks.
While Gemini 3 has garnered acclaim from tech luminaries like Sam Altman and Marc Benioff, Anthropic asserts that testers find Opus 4.5 more capable and autonomous. The model reportedly solves complex bugs without explicit guidance and tackles problems beyond the reach of its predecessors. The launch coincides with additional updates, though specifics remain undisclosed.
