Anthropic has unveiled its latest AI model, CLAUDE OPUS 5, which is being hailed as one of the top performers in the field. This model is nearly on par with Fable 5 but comes at half the price. In tests focused on programming and cognitive tasks, it has emerged as a new benchmark. Here are a few notable achievements of Opus 5: - When tasked with producing a 3D model of a machine part based solely on a provided blueprint—without direct access to the drawing—Opus 5 created its own computer vision pipeline to extract geometry from raw pixels, successfully reconstructing the entire machine component. It accomplished this repeatedly, while no competing model with a similar setup could solve the problem after five attempts. - Opus 5 detected a flaw in a widely used messaging application, identified the root cause, and rectified a specific issue that had been overlooked in a community-developed patch. In contrast, a rival model only addressed a surface-level symptom, failing to tackle the underlying problem before claiming to have fixed the error. - An engineer from a trading firm utilized Opus 5 to generate a stream of market data for a new exchange in just one session. Previous models were unable to handle this task, even with comprehensive guidance from the engineer. Lacking actual data streams for validation, Opus 5 even created its own testing environment to ensure that its code accurately processed the exchange data. How impressive is that, China?
Informational material. 18+.
