Anthropic launched its new artificial intelligence (AI) model, Claude Opus 5.5, on Tuesday. The company claims that this model performs on par with its predecessor, Claude 5.1, yet the cost of running it is 40 percent lower.
The launch of Claude Opus 5.5 comes at a time when debates regarding AI-related risks are intensifying. Anthropic stated that prior to the launch, Opus 5.5 underwent testing by independent AI safety research groups—Frontier Design and METR. According to Anthropic, Opus 5.5 outperformed its rival OpenAI’s GPT-5.6 model on software development benchmarks.
Opus 5.5 is priced at $4 per million input tokens and $20 per million output tokens, representing a 20 percent reduction compared to Claude 5. “Per million input tokens” refers to the cost incurred for every one million token segments (units of text) processed, while “per million output tokens” refers to the cost of generating one million tokens. The model is available on platforms such as Amazon Web Services, Google Cloud, and Microsoft Azure. The company reported that the model performed better than previous systems in internal safety tests.
