Anthropic releases Claude Opus 4.5, as AI war heats up
The past few weeks have seen new, impressive AI models debut from OpenAI and Google. Today it’s Anthropic’s turn to flex, as it releases Claude Opus 4.5, the latest iteration of its flagship AI model.
Anthropic’s Claude model is widely considered to be among the best at coding, and this model helps the company stay at the head of the pack.
Benchmarks released by Anthropic show Opus 4.5 besting both GPT 5.1 and Gemini 3 with an all-time high score of 80%, and the widely used “SWE-bench” coding benchmark. It also posted high scores for benchmarks measuring computer use, and the notoriously challenging ARC-AGI-2 visual problem-solving test, though apparently it can’t run a vending machine as profitably as Google’s Gemini 3 can.
AI coding is one of the few bright spots as companies seek profitable enterprise applications for AI that actually improve productivity. Anthropic’s success with enterprise customers has helped push its valuation to nearly $350 billion.
Benchmarks released by Anthropic show Opus 4.5 besting both GPT 5.1 and Gemini 3 with an all-time high score of 80%, and the widely used “SWE-bench” coding benchmark. It also posted high scores for benchmarks measuring computer use, and the notoriously challenging ARC-AGI-2 visual problem-solving test, though apparently it can’t run a vending machine as profitably as Google’s Gemini 3 can.
AI coding is one of the few bright spots as companies seek profitable enterprise applications for AI that actually improve productivity. Anthropic’s success with enterprise customers has helped push its valuation to nearly $350 billion.
