Anthropic has been experiencing some momentum lately. Following high-profile announcements like its collaboration with Amazon so that Claude helps power Amazon's recently previewed Alexa+ and the Claude 3.7 Sonnet and Claude Code launch, the company announced Monday it has closed a $3.5 billion Series E funding round at a $61.5 billion post-money valuation. Crunchbase reports that Anthropic has raised $18.2 billion after this funding round, led by Lightspeed Venture Partners. The Series E also saw participation from Bessemer Venture Partners, Jane Street, Menlo Ventures, and Salesforce Ventures, as well as new and existing investors.

News of this investment also comes at a time when not only the industry continues to experience the aftermath of DeepSeek R1's release, but the major AI companies are ramping up the development of the test-time compute techniques that endow LLMs with "reasoning" capabilities. This is because "reasoning" models have been long regarded as a key ingredient of the much anticipated AI agents. Although many details remain imprecise, AI agents are expected to be able to perform more meaningful work with less supervision.

How 'meaningful' this work is usually depends on the context, ranging from the level of a personal assistant that can accomplish tasks for its user such as booking restaurant and travel reservations or doing some relatively low-stakes shopping (OpenAI Operator, Alexa+, to name a few), to doing high-level work which includes assisting on 'PhD-level' scientific research (Google's co-scientist, OpenAI, reportedly) or performing the work equivalent to that performed by a mid-level software engineer (OpenAI, Meta).

Anthropic has differentiated itself in that it mostly makes more moderate claims than its rivals, and generally avoids marketing itself with the flashy and over-ambitious claims that characterize many of its competitors' strategies. Still, the company stated in the Series E announcement that it "is focused on developing AI systems that can serve as true collaborators, working alongside teams to tackle complex projects, synthesize information across fields, and help organizations achieve outsized impact."