Anthropic eyes $2T valuation as quarterly revenue hits $11.5B

AI developer Anthropic is reportedly preparing for an initial public offering in October with an ambitious valuation target of $2 trillion. The target follows a massive financial surge, with the company's preliminary quarterly revenue climbing to $11.5 billion, representing a 14x increase from the $787 million recorded during the same period last year. Looking ahead, the artificial intelligence lab projects its annual revenue will scale to between $190 billion and $200 billion by the year 2028.

This rapid growth comes amid broader concerns over the immense capital required to sustain the artificial intelligence boom. A recent analysis reveals that nine leading technology firms are currently carrying $3 trillion in off-balance-sheet commitments. This hidden spending includes $1.2 trillion in data center leases alongside $1.9 trillion in hardware obligations. The massive scale of these financial arrangements has already dragged major industry players like Alphabet, Amazon, and Meta into negative free cash flow, making it increasingly difficult for investors to evaluate total corporate liabilities.

For software developers and enterprise practitioners, these staggering financial figures highlight the intense pressure on AI providers to monetize their tools quickly. While massive infrastructure investments ensure that developers have access to increasingly powerful models, they also signal potential volatility in API pricing and platform stability as startups and tech giants scramble to justify their capital expenditures. Practitioners must weigh the benefits of cutting-edge models against the long-term viability of providers navigating these extreme cash-burn rates.

Meanwhile, smaller open-source ecosystems continue to advance desktop-level capabilities. Nous Research has launched a new Bot Mode for its Hermes Desktop application, which is compatible with macOS, Windows, and Linux. This update assigns individual models, skillsets, and memories to local agents, allowing them to communicate with each other and share context across different message threads to streamline local workflows.