Anthropic releases Claude Opus 4.7, a ‘less capable’ model than Mythos

Anthropic announced a new artificial intelligence model on Thursday. Claude Opus 4.7The company said it’s an improvement over past models but “less capable overall” than its latest offering, the Claude Mythos Preview.

Anthropic said Claude Opus 4.7 is better at software engineering, following instructions, completing real-world tasks, and using file system-based memory. However, the model’s cyber capabilities are not as advanced as the Claude Mythos Preview, which Anthropic rolled out to select companies earlier this month as part of a new cybersecurity initiative called Project Glasswing.

“We are releasing Opus 4.7 with security measures that automatically detect and block requests that indicate prohibited or high-risk cybersecurity uses,” Anthropic said in a statement. he said. “What we learn from the real-world application of these security measures will help us work towards our ultimate goal of wide release of Mythos-class models.”

Anthropic said it tested efforts to “differentially reduce” Claude Opus 4.7’s cyber capabilities during training. The company encouraged security professionals who want to use the model for “legitimate cybersecurity purposes” to apply through a formal verification program.