Overview of Opus 4.5
Anthropic has unveiled its latest AI model, Opus 4.5, completing the 4.5 series. This model showcases advanced capabilities across various benchmarks, including coding and problem-solving tasks. Opus 4.5 stands out as the first to achieve over 80% on the respected SWE-Bench coding benchmark, highlighting its exceptional performance in software engineering.
Key Features and Improvements
- Opus 4.5 excels in multiple benchmarks like SWE-Bench and ARC-AGI 2.
- The model introduces enhanced computer use and spreadsheet functionalities.
- Anthropic has launched Claude for Chrome and Claude for Excel, expanding access to these tools.
- Memory management improvements allow for better handling of long-context operations and an “endless chat” feature for paid users.
Significance of the Launch
The release of Opus 4.5 is significant as it not only improves AI capabilities but also focuses on practical applications. The model is designed for agentic use cases, where it can manage tasks effectively as a lead agent. The memory enhancements are crucial for navigating complex tasks, such as exploring codebases and large documents. With strong competition from models like OpenAI’s GPT 5.1 and Google’s Gemini 3, Opus 4.5 aims to solidify Anthropic’s position in the AI landscape.
