Anthropic Tipped to Launch the Claude Opus 4.5 AI Model This Week

Anthropic could release the frontier model of its Claude 4.5 family soon. As per the tipster, the Claude Opus 4.5 artificial intelligence (AI) model was spotted on another platform, while the model's release table was also spotted separately. Based on the leaks, it is said that the large language model could arrive on Monday, bringing improvement across various parameters. Not a lot is known about the AI model at present, but based on the improvements in Claude 4.5 Sonnet and Claude 4.5 Haiku, coding performance and agentic capabilities could be a major focus area.

Anthropic Could Release Claude Opus 4.5 Soon

Tipster @kimmonismus on X (formerly known as Twitter), claimed in a post that Anthropic's Claude Opus 4.5 was being readied for release and it was spotted on Poe, an AI chatbot platform that hosts third-party models. Later, in a separate post, the tipster shared a release table of the LLM, claiming it is scheduled to be released on Monday, November 24.

Social media on Sunday wa abuzz with rumours about the release of Anthropic's frontier model. Founder of Zenjoy, Peter Dedene, also shared a screenshot of the release table that mentions that “Claude Kayak” is releasing on Monday. It is said that Kayak is an internal nickname for Opus 4.5.

Apart from leaks about its release date, no other reliable information about the AI models is currently known. It has been speculated that the AI model will further improve code generation and bring greater agentic capabilities to users; however, upgrades in other areas are not known.

Notably, Anthropic released Claude 4.5 Sonnet in September and Claude 4.5 Haiku, the fastest model in the family, arrived in October. These models displayed improvements in coding, agentic operations, computer use, reasoning, and domain-specific knowledge. Sonnet also managed a score of 77.2 percent on the SWE bench-Verified benchmark, which tests AI models on their coding capabilities. This score was higher than what OpenAI's GPT-5 and Google's Gemini 2.5 achieved.

Meanwhile, recently, Claude was used to conduct a large-scale agentic cyberattack. In a detailed analysis, Anthropic claimed it to be first-of-its-kind incident where the AI carried out most of the hacking, with minimal input from a human operator.