Anthropic has launched Anthropic’s Opus 4.5, the latest and final version of its 4.5 AI series. The release follows Sonnet 4.5 in September and Haiku 4.5 in October. The new model delivers advanced performance across coding, tool use, and reasoning tasks.
Anthropic’s Opus 4.5 leads on multiple industry benchmarks, including SWE-Bench, Terminal-bench, tau2-bench, MCP Atlas, ARC-AGI 2, and GPQA Diamond. Notably, it is the first AI to score above 80% on SWE-Bench, a key measure of coding reliability. This makes it capable of outperforming human software engineers in coding tasks.
The model includes major improvements in computer interaction and spreadsheet handling. To support this, Anthropic is expanding access to its Claude tools for Chrome and Excel. The Chrome extension is now available to all Max users, while the Excel model will be accessible to Max, Team, and Enterprise users.
Anthropic’s Opus 4.5 also features memory and long-context enhancements. Dianne Na Penn, head of product management for research at Anthropic, explained that improving context handling requires not just larger windows but smarter memory retention. These upgrades allow the AI to compress older conversation data automatically, enabling an “endless chat” feature for paying Claude users.
Many improvements focus on agent-based workflows, where Anthropic’s Opus 4.5 coordinates multiple Haiku-powered sub-agents. Strong memory helps the AI navigate codebases, review long documents, and manage complex tasks effectively.
In other related news also read Best Software Engineering Universities In Pakistan
The new model enters a competitive market alongside OpenAI’s GPT 5.1 and Google’s Gemini 3. With these updates, Anthropic positions Opus 4.5 as a leading AI in coding, multi-agent workflows, and long-context reasoning.
