Blab AI
Enterprise AI Spend Tightens as Firms Shift From OpenAI and Anthropic to Cheaper Alternatives
When the AI boom that began in 2022 suddenly turned into a budgeting nightmare, many large‑scale business customers started cutting back on the high‑priced services of OpenAI and Anthropic. According to a CNBC report dated 26 June 2026, firms are now gravitating toward lower‑cost competitors such as DeepSeek.
The shift follows a period of rapid AI adoption that began with OpenAI’s ChatGPT, which attracted corporate interest across industries. Enterprise teams, especially those focused on AI‑assisted coding, pushed large volumes of tokens into the models without closely monitoring expenses. The result was a surge in spending that many companies now find unsustainable.
Flo Crivello, chief executive of AI startup Lindy, told CNBC that the firm moved 100 % of its traffic from Anthropic’s Claude models to DeepSeek’s open‑weight alternatives. Crivello said the change would save Lindy millions of dollars within months and described the decision as a matter of survival for the business.
Other enterprises are taking similar steps. Uber, which reportedly burned through its entire annual AI budget in just four months, has introduced tiered pricing for some AI tools, with the lowest tier starting at $1,500 per month. Jeff Henry, president of consulting at Highspring, said that some of his clients are pulling back until they can prove a return on investment, while others are waiting 12 to 18 months before committing to major spending.
The timing of the shift coincides with confidential initial public offering (IPO) filings by OpenAI and Anthropic. Both companies filed S‑1 prospectuses with the Securities and Exchange Commission in early June. Anthropic’s annualized revenue run rate reached $47 billion in May, while OpenAI’s was near $25 billion earlier in the year, according to the report.
Gil Luria, an equity analyst at D.A. Davidson, suggested that the filings may be deliberate. “There has to be some period of time in the future where there’s some rationalizing of spend by companies, and that may be a blip ahead for Anthropic and OpenAI,” Luria said. “That creates some sense of urgency to go public before we see that.”
The budget pressure is opening space for competitors. Microsoft, Amazon, and Google have each expanded lower‑cost model offerings aimed at business customers. OpenAI announced significant price cuts to its services and introduced new spending analytics and controls for enterprise administrators earlier this month. Anthropic rolled out similar controls that allow organizations to set spending limits at the individual and organizational level.
Neither OpenAI nor Anthropic responded to requests for comment.
The move reflects a broader trend of enterprises seeking clearer returns on AI investment. With the cost of large‑language‑model (LLM) usage rising, many firms are re‑evaluating the value delivered by premium models. The availability of cheaper, open‑weight alternatives such as DeepSeek, which has positioned its models under free and open‑source licenses, offers a lower‑barrier option for companies that need high‑performance language models without the premium price.
In the coming months, the industry will likely see continued experimentation with cost‑effective models and tighter spending controls. Companies that have already reduced or redirected their AI budgets may keep pace with new pricing structures from major vendors, while smaller enterprises may adopt open‑source solutions to maintain AI capabilities.
The current situation is that enterprise AI spend is tightening, leading to a shift away from OpenAI and Anthropic toward cheaper alternatives. OpenAI and Anthropic are preparing for public listings, and both have introduced spending controls. The market will continue to evolve as businesses seek clearer ROI and as competitors expand lower‑cost offerings.
The shift follows a period of rapid AI adoption that began with OpenAI’s ChatGPT, which attracted corporate interest across industries. Enterprise teams, especially those focused on AI‑assisted coding, pushed large volumes of tokens into the models without closely monitoring expenses. The result was a surge in spending that many companies now find unsustainable.
Flo Crivello, chief executive of AI startup Lindy, told CNBC that the firm moved 100 % of its traffic from Anthropic’s Claude models to DeepSeek’s open‑weight alternatives. Crivello said the change would save Lindy millions of dollars within months and described the decision as a matter of survival for the business.
Other enterprises are taking similar steps. Uber, which reportedly burned through its entire annual AI budget in just four months, has introduced tiered pricing for some AI tools, with the lowest tier starting at $1,500 per month. Jeff Henry, president of consulting at Highspring, said that some of his clients are pulling back until they can prove a return on investment, while others are waiting 12 to 18 months before committing to major spending.
The timing of the shift coincides with confidential initial public offering (IPO) filings by OpenAI and Anthropic. Both companies filed S‑1 prospectuses with the Securities and Exchange Commission in early June. Anthropic’s annualized revenue run rate reached $47 billion in May, while OpenAI’s was near $25 billion earlier in the year, according to the report.
Gil Luria, an equity analyst at D.A. Davidson, suggested that the filings may be deliberate. “There has to be some period of time in the future where there’s some rationalizing of spend by companies, and that may be a blip ahead for Anthropic and OpenAI,” Luria said. “That creates some sense of urgency to go public before we see that.”
The budget pressure is opening space for competitors. Microsoft, Amazon, and Google have each expanded lower‑cost model offerings aimed at business customers. OpenAI announced significant price cuts to its services and introduced new spending analytics and controls for enterprise administrators earlier this month. Anthropic rolled out similar controls that allow organizations to set spending limits at the individual and organizational level.
Neither OpenAI nor Anthropic responded to requests for comment.
The move reflects a broader trend of enterprises seeking clearer returns on AI investment. With the cost of large‑language‑model (LLM) usage rising, many firms are re‑evaluating the value delivered by premium models. The availability of cheaper, open‑weight alternatives such as DeepSeek, which has positioned its models under free and open‑source licenses, offers a lower‑barrier option for companies that need high‑performance language models without the premium price.
In the coming months, the industry will likely see continued experimentation with cost‑effective models and tighter spending controls. Companies that have already reduced or redirected their AI budgets may keep pace with new pricing structures from major vendors, while smaller enterprises may adopt open‑source solutions to maintain AI capabilities.
The current situation is that enterprise AI spend is tightening, leading to a shift away from OpenAI and Anthropic toward cheaper alternatives. OpenAI and Anthropic are preparing for public listings, and both have introduced spending controls. The market will continue to evolve as businesses seek clearer ROI and as competitors expand lower‑cost offerings.
