Anthropic on Tuesday announced the release of Claude Sonnet 4.6, the second major update to its artificial intelligence (AI) suite in just 12 days.
The latest version, which becomes the default for both free and Pro users, represents a full upgrade in the model’s ability to handle complex coding, agent planning, and autonomous computer use, the company said.
The launch underscores the breakneck pace of the generative AI industry as Anthropic, OpenAI, and Google battle for market dominance. By bringing Opus-class performance to its mid-tier Sonnet model, Anthropic is effectively commoditizing high-end reasoning capabilities for a broader user base.
The standout feature of Sonnet 4.6 is a significantly expanded long-context window. The model can now process up to 1 million tokens in a single request, double the previous capacity for the Sonnet line. According to Anthropic, this allows the AI to ingest entire software codebases, dozens of research papers, or massive legal contracts simultaneously.
Anthropic admitted the model “certainly still lags behind the most skilled humans at using computers,” but nonetheless described its trajectory as “remarkable.”
The model achieved record benchmark scores on OS World for computer use and scored 60.4% on ARC-AGI-2, a metric designed to measure human-like reasoning. The figures place Sonnet 4.6 ahead of most peers, though it continues to trail flagship models like its sibling Opus 4.6 and Google’s Gemini 3 Deep Think.
Rapid deployment of AI tools is beginning to rattle global financial markets. Anthropic’s recent focus on its Claude Cowork agent, which is capable of automating tasks in legal, product management, and data analysis, has fueled investor anxiety regarding the displacement of traditional software services.
Ripple effects are evident across the globe. The iShares Expanded Tech-Software Sector ETF (IGV) has plummeted more than 20% so far this year. Meanwhile, major IT firms in India and Japan’s Nomura Research have seen shares slide, with the latter hitting a 52-week low.
At the same time, a Bank of America survey released Tuesday revealed that 25% of fund managers view “AI over-inflation” as a top market risk, while 30% fear a looming credit crisis driven by massive infrastructure spending.
Despite market jitters, the appetite for AI investment remains voracious. Anthropic confirmed on Thursday that it closed a $30 billion funding round, skyrocketing its valuation to $380 billion — twice its value in late 2025. Capital injection comes as industry titans including Alphabet, Amazon.com Inc., Meta Platforms Inc., and Microsoft Corp. project a combined $610 billion in AI-related capital expenditures for 2026.
Founded in 2021 by former OpenAI executives, Anthropic has positioned itself as a safety-conscious alternative to its rivals. However, with Sonnet 4.6 now live and a refreshed Haiku model expected in the coming weeks, the company is proving it can match the aggressive shipping cycles of its most well-funded competitors.
