Hajra Naz
Anthropic announced Opus 4.5 on Monday. This is the newest version of its flagship model. It also completes the 4.5 series. Sonnet 4.5 launched in September. Haiku 4.5 launched in October.
Opus 4.5 delivers strong benchmark performance. It scores high in coding tests like SWE-Bench and Terminal-bench. It performs well in tool-use tests such as Tau2-Bench and MCP Atlas. It also shows strong results in general problem-solving with ARC-AGI 2 and GPQA Diamond.
Opus 4.5 is the first model to score above 80% on the SWE-Bench verified. This is a respected coding benchmark used across AI research and developer workflows.
Anthropic highlighted Opus’s computer-use skills and spreadsheet handling. The company released new products to show how Opus 4.5 performs in real work settings. Claude for Chrome and Claude for Excel will now roll out widely.
The Chrome extension will be available to all Max users. The Excel-focused version will reach Max, Team, and Enterprise users. These additions enhance productivity and drive AI automation trends across the web and workplace applications.
Opus 4.5 also includes major memory upgrades. These changes improve long-context tasks and require deep adjustments to the model’s internal memory system.
Dianne Na Penn, Anthropic’s head of product management for research, told TechCrunch that long context alone is not enough. She said the model must also identify the right details to save and recall during complex tasks.
These improvements also enabled an “endless chat” feature for paid Claude users. Chats will continue even when Opus reaches its context limit. The model will compress its memory silently without interrupting the user.
Many of these updates support agentic AI use cases. Opus 4.5 can act as a lead agent. It can manage groups of Haiku-powered sub-agents. This requires strong working memory and precise control over large code bases and long documents. Penn said this is where the new memory upgrades become essential.
Opus 4.5 now enters a competitive field. It will compete with OpenAI’s GPT 5.1, which launched on November 12, and Google’s Gemini 3, released on November 18. These models are shaping current AI trends across automation, coding AI, and advanced AI assistants.
FAQs
1. What makes Opus 4.5 different from earlier versions?
Opus 4.5 brings stronger benchmark scores, improved memory systems, and better performance on coding and tool-use tasks.
2. Is the Chrome extension available to all users?
Yes. Claude for Chrome will be available to all Max users after the wider rollout.
3. What is the “endless chat” feature?
It allows chats to continue without interruption. Opus compresses the older context automatically and keeps the conversation active.
4. How does Opus 4.5 handle agentic AI tasks?
It can act as a lead agent and coordinate Haiku-based sub-agents. Strong memory upgrades support these scenarios.
5. How does Opus 4.5 compare to GPT 5.1 and Gemini 3?
It competes directly with both models in coding, general reasoning, and AI-assistant automation, forming part of the current frontier-model race.
