AI Pricing Challenges Grow as Businesses Struggle to Control Rising Token Costs
Companies adopting AI agents and large language models face growing uncertainty over pricing as unpredictable token usage complicates budgeting and long-term planning.
Businesses are rethinking AI spending as token usage continues to surge.
AI Pricing has emerged as one of the biggest challenges facing businesses as companies invest heavily in artificial intelligence while struggling to predict the cost of using large language models (LLMs) and AI-powered agents.
Technology giants including Microsoft, Google and Anthropic have invested hundreds of billions of dollars in developing AI models such as ChatGPT, Gemini and Claude. While many consumers continue to use free versions of these tools, businesses increasingly rely on paid services that offer advanced capabilities for coding, automation and enterprise applications.
More than 100 words into the article – AI Pricing industry experts say the rapid growth of AI agents has made pricing more complicated because costs depend on token consumption, which varies significantly from one task to another. Tokens are the small units of data that AI models process to understand prompts and generate responses. Since prompts and outputs differ in length and complexity, businesses often struggle to predict their monthly AI expenses.
Token Costs Remain Difficult to Predict
Simon Gooch of identity management company Saviynt said setting long-term pricing models has become increasingly difficult because AI economics continue to evolve.
He noted that companies cannot accurately estimate costs over the next several years as token pricing and AI usage patterns continue to change.
According to Goldman Sachs, monthly token consumption could increase 24-fold between 2026 and 2030, reaching 120 quadrillion tokens as businesses adopt more AI agents.
Michael Smith Named Pakistan Batting Coach Ahead of England Test Series
Businesses Face Budget Challenges
Many organisations underestimate how quickly they consume AI tokens until they receive unexpectedly high monthly bills.
Reports suggest that even major companies such as Microsoft have limited employees’ use of certain third-party AI coding tools, while Uber reportedly exhausted its annual AI coding budget within a few months.
Will Venters, Associate Professor of Digital Innovation and Information Systems at the London School of Economics, said businesses often struggle because AI produces non-deterministic outputs, making both costs and value difficult to forecast.
Companies Explore Alternative Pricing Models
Some businesses continue using consumer AI subscriptions for limited commercial work to reduce expenses, although industry experts believe large AI providers may eventually tighten restrictions as they seek stronger profitability.
Software executives also recommend writing more detailed prompts and selecting the most appropriate AI model for each task to reduce unnecessary token consumption.
AI Providers Yet to Find a Standard Model
As more companies integrate AI into products and services, pricing becomes increasingly difficult to manage. Costs can rise quickly as organisations deploy AI for software development, testing, cybersecurity and automated decision-making.
Bill Peterson, Senior Director of Product Marketing at Sumo Logic, said software providers are still debating how to charge customers for AI-powered services.
Companies are considering several approaches, including subscription price increases, usage-based billing and bundled pricing models. However, frequent pricing changes by AI platform providers continue to make budgeting difficult for enterprise customers.
Industry experts say businesses are likely to refine their pricing strategies as AI adoption accelerates, but a widely accepted commercial model has yet to emerge.
