Anthropic launches the new Claude Opus 4.7 model, enhancing coding and multimodal capabilities. Anthropic has introduced the new Claude Opus 4.7 model, with a focus on improving performance in complex software engineering and long-form tasks. According to the official announcement, the model outperforms Opus 4.6 in automated planning, identifying and correcting its own logical errors, and handling complex asynchronous and CI/CD workflows. It achieves approximately a 13% increase in task completion rates across 93 coding benchmarks conducted by partner organizations and solves more real-world production tasks on benchmarks such as Rakuten-SWE-Bench. Opus 4.7 also enhances high-resolution image understanding, supporting images of up to approximately 3.75 million pixels for multimodal use cases including code review, document analysis, and life sciences patents. Pricing remains unchanged from Opus 4.6, with the API model named “claude-opus-4-7,” and new control options including “xhigh” inference intensity and task budget settings have been added.