At a Glance · AI-generated

Anthropic released Claude Opus 4.7 on April 16, advancing the model's coding and visual reasoning capabilities while implementing automated cybersecurity safeguards that deliberately limit potentially dangerous cyber abilities. The new model positioned itself as a safer alternative amid anticipation for Anthropic's Mythos preview. By April 19, analysis revealed that Opus 4.7 consumes significantly more tokens than its predecessor despite Anthropic's flat pricing structure, raising questions about the model's cost efficiency.