Claude Opus 5 (Max)

Claude Opus 5 is a large language model developed by Anthropic, released on July 24, 2026. Positioned within the Claude model family, it is designed to approach the frontier intelligence of Claude Fable 5 while operating at roughly half the price, and it succeeds Claude Opus 4.8 as Anthropic's flagship Opus-tier model.

Capabilities and Performance

Opus 5 is built for coding, agentic workflows, and general knowledge work. On internal evaluations such as Frontier-Bench and GDPval-AA, Anthropic reported that Opus 5 sets a new state-of-the-art among its models, though it remains behind Mythos 5 on cybersecurity tasks. On Frontier-Bench v0.1, the company said the model surpasses all other Anthropic models and more than doubles Opus 4.8's performance at a lower cost per task, while at maximum effort it performs within roughly 0.5 percent of Fable 5's peak score at half the cost per task. On ARC-AGI 3, an evaluation involving novel problem-solving, its score is reported as three times that of the next-best model.

Design and Controls

The model includes an adjustable "effort" setting, allowing users to select low, medium, or high effort levels to balance cost, speed, and capability for a given task. It is also offered in a Fast mode, which runs at around 2.5 times the default speed at twice the base price. Alongside the release, Anthropic introduced two beta features: mid-conversation tool changes that let developers alter which tools the model can use without invalidating the prompt cache, and automatic fallbacks that route flagged requests to another model rather than blocking them.

Safety

Anthropic has described Opus 5 as the most aligned Opus model to date and the least susceptible to being tricked into misuse. On biology-related safeguards, Opus 5 uses a similar safety suite to Opus 4.8, making it Anthropic's most capable generally available model for scientific research, though the company noted it still shows limitations on long-running, autonomous research tasks, an area where Mythos 5 remains the stronger model.