Anthropic reveals new Opus 4.7 model with focus on advanced software engineering

Key Points:

• Anthropic has released Claude Opus 4.7, the latest version of its AI model focused on advanced software development, arriving two months after Opus 4.6 in line with the company’s bi-monthly upgrade schedule.
• Opus 4.7 demonstrates significant improvements in handling complex coding tasks with less supervision, enhanced instruction following, multimodal support, and better memory for long, multi-session work.
• The model features higher-resolution image processing and produces higher-quality creative outputs such as interfaces, slides, and documents, outperforming Opus 4.6, GPT-5.4, and Gemini 3.1 Pro but still trailing behind the more capable Claude Mythos Preview.
• Anthropic updated the tokenizer in Opus 4.7, leading to increased token usage per input and more output tokens due to deeper reasoning on difficult problems, though efficiency gains were noted by users like Box, which reported faster responses and reduced resource use.
• Alongside the model update, Anthropic enhanced Claude Code in the Claude Mac app, introducing “auto mode” for Max plan subscribers and a new “/ultrareview” command to improve code review safety and thoroughness.