To help you understand the trends surrounding business and technology and what we expect to happen in the future, our highly experienced Kiplinger Letter team will keep you abreast of the latest developments and forecasts. (Get a free issue of The Kiplinger Letter or subscribe.) You'll get all the latest news first by subscribing, but we will publish many (but not all) of the forecasts a few days afterward online. Here's the latest…

A pricing war is afoot among cutting-edge artificial intelligence vendors. Meta will jolt the competition with cheaper advanced AI tools for businesses from its new AI model, Muse. SpaceX’s latest Grok AI is built for efficiency and low costs, priced 60% cheaper than Anthropic. Microsoft is shifting from Anthropic and OpenAI to its own, cheaper internal AI tools for apps such as Excel and Outlook.

Meanwhile, cheaper Chinese AI models such as DeepSeek and Kimi are quickly gaining ground, though congressional investigations may lead to attempts to put curbs on the foreign tech. The latest version of Kimi, developed by Chinese company Moonshot AI, has sparked equal measures of excitement from U.S. customers and concern from leading American AI companies and federal officials. U.S. policy concerns include China’s massive government subsidies, intellectual property theft, cybersecurity risks and the general threat of Chinese competition.

The competition could put pressure on profit margins for AI leaders, which still must invest massive sums of money to develop and deploy leading tech. The latest Grok 4.5 model is a threat to Anthropic and OpenAI since it is a "'good enough,' fast, and super-cheap model," writes Neil Shah, analyst at Counterpoint Research, in a recent post. “Now enterprises have an attractive option, allowing them to optimize their AI spend before it spirals out of control.”

But there’s a catch for companies excited to see lower prices: Customers aren’t likely to save money because their AI use is rising so fast — AI is billed based on how much is consumed, which is far outpacing per-unit cost declines.

That remains true even as the long-term trends look promising for customers. Market research firm Gartner says a combination of efficiency improvements in chips, data centers, software and more will drive down prices. "By 2030, performing inference on a large language model with one trillion parameters will cost GenAI providers over 90% less than it did in 2025," according to their analysis from March.

Not all the savings will be passed on to customers, says Gartner, and cutting-edge agentic AI, which automates all sorts of computing tasks, consumes far more AI compute. "Agentic models, for example, require between 5-30 times more tokens per task than a standard generative AI chatbot," says Gartner. (Tokens are the units of data processed by AI. Companies are commonly billed by how many tokens they use.)

So what are companies going to do to reel in AI budgets? “Enterprises will learn and become prudent not to stick to one vendor or model,” according to Shah. That means that most complex tasks can be accomplished with the most expensive, best AI. Simpler tasks can be done with cheaper AI tools.

Gartner recommends that companies maintain a list of tasks that require high-end AI tools, train employees on how to reduce AI costs, post limits on individual token consumption and closely track AI usage.

This forecast first appeared in The Kiplinger Letter, which has been running since 1923 and is a collection of concise weekly forecasts on business and economic trends, as well as what to expect from Washington, to help you understand what’s coming up to make the most of your investments and your money. Subscribe to The Kiplinger Letter.

Related Content

• The Memory Crunch Wallops the Phone and PC Market
• Beyond AI: Why Our Top Dividend Stocks Remain Reliable Picks
• AI is Powering A Semiconductor Boom
• Artificial Intelligence is Raising Cyber Threats