Anthropic has launched its latest model, Claude Opus 4.8, with enhanced capabilities and the ability to provide more honest answers than its predecessor. Other improvements over Opus 4.7 include the ability to code agents, perform thought-provoking tasks, analyze financial data, and control hundreds of AI agents simultaneously for a given task.
Early testers reported that b4.8 was more likely to express uncertainty about its work and was less likely to make unsupported claims. In an assessment within Anthropic, Opus 4.8 was about four times less likely to allow defects in the code it had written.
Claude Opus 4.8 is available today at the same price as Opus 4.7. Anthropic is currently developing a model that is comparable to Opus 4.8 but at a lower cost.
