During the AI Impact Summit held in New Delhi on February 19, 2026, Indian Prime Minister Narendra Modi took part in a group photograph with notable leaders from the artificial intelligence sector, including OpenAI's CEO Sam Altman and Anthropic's CEO Dario Amodei.
In a recent shift aimed at controlling expenses, Flo Crivello, CEO of the AI startup Lindy, transitioned his company away from using Anthropic's Claude models, opting instead for the more affordable options provided by the Chinese company DeepSeek. This move has already shown promising signs, with Crivello noting a significant drop in costs. "We did it, and you could see that cost curve go down, like, crash to the ground," he remarked during an interview from Lindy's San Francisco location. He estimates that this change will save the company millions in the coming months, although spending on AI is expected to exceed payroll expenses. "It's a matter of survival for the business," he stated. Having previously worked at Uber, Crivello is part of a growing trend of U.S. executives looking to rein in escalating AI costs that have soared since the launch of OpenAI’s ChatGPT in 2022. This widespread adoption of AI technology for various business functions has led to skyrocketing expenses, particularly in AI-assisted coding, where developers have flooded the market with tokens for new tools that used to require large coding teams.
In response to these surging costs, Uber announced earlier this month that it is implementing spending tiers for its AI tools, starting at $1,500 per month, while allowing employees to request access to higher tiers. Just four months into the fiscal year, Uber's Chief Technology Officer Praveen Neppalli Naga disclosed that the company had already exceeded its entire yearly AI budget.
Both OpenAI and Anthropic have significantly benefited from the rush to spend, experiencing rapid growth that has pushed their valuations to nearly $1 trillion. However, as they prepare for potential IPOs—having filed confidentially in June—the landscape surrounding AI is shifting, with leaders like Crivello seeking tangible returns on their investments before increasing their expenditures. "Current growth rates for Anthropic and OpenAI are the fastest they will ever be, which is mostly a matter of basic math," said Gil Luria, an equity analyst with D.A. Davidson. With annualized revenue figures vastly higher than previous years, now may be a strategic time for these companies to go public, particularly as some clients begin to curb excessive token spending. Recent reports indicated that Anthropic is on track for a $47 billion revenue run rate, a substantial jump from its $10 billion earnings last year, while OpenAI's run rate was estimated near $25 billion.
Crivello, while appreciating Anthropic, acknowledged that his company has faced "unsustainable" AI costs for quite some time. Lindy was founded on the expectation that token costs would decline significantly—a trend that, while initially true, hasn't continued as anticipated. Crivello expressed willingness to revert to Claude models if prices were to drop. "I hope that they cut the costs again at some point but, until then, we've got options," he added. Jeff Henry from Highspring reported a similar sentiment among clients who are now holding back expenditures until they can establish a clear ROI, with many possibly waiting one to one and a half years before making new commitments in AI.
The sentiment was echoed by Darren Kimura, CEO of AISquared, who noted that companies are using advanced AI models for simpler tasks that could be performed by cheaper alternatives. This has led to the emergence of model routing techniques, but Kimura cautioned that dependency on high-end models would become unsustainable. Luria mentioned that the pricing environment remains "unsophisticated," but both leading AI firms are attempting to adapt to these budget-conscious trends. OpenAI recently introduced new analytics tools to help enterprises manage their spending better, while Anthropic also implemented controls for budget management at the organizational level.
As companies become increasingly cost-aware, both OpenAI and Anthropic face competition from technologically sophisticated rivals eager to establish lower-cost offerings. Microsoft has invested heavily in both companies and recently launched a line of budget-friendly models while emphasizing their GitHub Copilot's ability to recommend the most suitable models for users. Amazon’s AI head indicated plans to compete with top-tier models in the years ahead, and Google used its recent developer conference to showcase cost-effective AI solutions. According to Sundar Pichai, Google introduced the Gemini 3.5 Flash, which provides capabilities at significantly lower prices compared to leading models.
Concerning their IPO timelines, neither OpenAI nor Anthropic have disclosed specific plans, but reports suggest OpenAI might delay its public offering until next year. As these firms navigate intense competition and a changing capital landscape, pressure to go public may stem from their financial requirements, as traditional funding sources become limited. "A lot of the traditional pockets of capital are drying up," said Dharmesh Thakker of Battery Ventures, highlighting the need for fresh investment avenues to support their ambitious growth.
