The AI sector continues its rapid pace of advancement and development. Every so often we hear news about a new AI model or an update to a renowned LLM. Today we have the latter case as Anthropic has launched Claude Opus 4.5, its latest large language model.
Claude Opus 4.5 is the third significant update from the company in two months. The previous ones were Sonnet 4.5 and Haiku 4.5. Anthropic has always proudly boasted about how powerful Claude is in assisting professionals—especially developers and various knowledge workers. The latest update enhances these capabilities to handle very complex tasks.
Claude Opus 4.5, Anthropic’s latest AI model update
The model demonstrates particular strengths in agentic coding and computer control. It can effectively resort to several tools to complete multi-step workflows. Opus 4.5 is engineered for efficiency in these tasks; Anthropic notes that agents built on the model can reach their highest performance level after only four attempts at task refinement. This represents a notable reduction compared to other LLMs that may require ten.
To illustrate its capabilities, Anthropic cited the model’s performance on standard and internal evaluations. On the SWE-bench Verified test set, which assesses real-world software engineering abilities, Claude Opus 4.5 reportedly achieved scores surpassing those of competing models, including Google’s Gemini 3 Pro and OpenAI’s GPT-5.1.
The company also tested the model on a complex, internal performance engineering exam. Opus 4.5 scored higher than any human candidate previously tested under the same two-hour constraints. These results reflect the model’s ability to apply technical judgment under pressure. That said, the company notes the test does not measure intuitive or social skills.
Practical utility and pricing adjustments
Beyond core performance metrics, Opus 4.5 introduces features designed to enhance enterprise productivity. The model’s reasoning capabilities are integrated directly into key workplace applications. Claude for Excel, which handles spreadsheet tasks, pivot tables, and chart generation, is now widely available to Max, Team, and Enterprise users. Additionally, Claude Code, the dedicated programming assistant, is now accessible through Anthropic’s desktop app.
Lastly, using the latest model is cheaper than ever. Anthropic reduced the cost for Opus 4.5 input tokens by approximately two-thirds compared to its predecessor, Opus 4.1. This tweak helps maintain a competitive position in this key market.
Opus 4.5 is now available across all Anthropic services and developments. This includes their apps, APIs, and even major cloud platforms, so everyone can access the capabilities of their latest AI model.
