Why Firms Are Struggling to Set Prices for AI Services
Firms like Microsoft, Google, and Anthropic are investing heavily in AI, but setting prices for AI services is proving difficult due to rapidly changing economics around tokens, the building blocks of LLMs and agentic AI.
Intelligence analysis by Llama
Companies are struggling to manage the costs of AI services, which are unpredictable and non-deterministic. This is making it difficult for them to set prices for their services, and some are finding ways to work around this by using flat fee personal accounts or being more precise with their prompts.
Imagine you have a magic machine that can do lots of things for you, like write stories or make music. But the machine uses special tokens to work, and it's hard to know how many tokens it will use. This makes it hard for companies to charge people for using the machine, because they don't know how much it will cost. It's like trying to buy a toy without knowing how much it will cost, or how long it will take to make it.
Analysis
The Unpredictable Economics of AI Tokens
The cost of individual tokens, or the credits used to pay for them, has plummeted in recent years. However, the number of tokens consumed by businesses and consumers has skyrocketed. According to analysis by Goldman Sachs, external token consumption will increase 24 times between 2026 and 2030 to 120 quadrillion tokens a month.
This has created a challenge for companies trying to set prices for their AI services. The cost of these services is not entirely predictable, and subtle variations in the prompt can produce different answers. Different models will produce different answers, and the same prompt will not always produce the same answer.
The Rise of Agentic AI
Agentic systems use multiple AI agents together to make decisions and take actions, further increasing both token use and unpredictability. While the cost of individual tokens has plummeted, the number of tokens consumed by businesses and consumers has skyrocketed.
The Challenge of Managing AI Costs
Companies are finding it difficult to manage the costs of AI services. Staff are burning through tokens, and companies are struggling to keep track of their token usage. This is making it difficult for them to set prices for their services, and some are finding ways to work around this by using flat fee personal accounts or being more precise with their prompts.
The Future of AI Pricing
The struggle to set prices for AI services has significant implications for the development and implementation of AI technology. Companies will need to find ways to manage the costs of AI services, and set prices that reflect the value of these services. This will require a better understanding of the economics of AI tokens, and the development of new pricing models that take into account the unpredictability of AI costs.
Key points
- Firms are struggling to set prices for AI services due to rapidly changing economics around tokens.
- The cost of individual tokens has plummeted, but the number of tokens consumed by businesses and consumers has skyrocketed.
- Agentic systems use multiple AI agents together to make decisions and take actions, further increasing both token use and unpredictability.
- Companies are finding it difficult to manage the costs of AI services, and are struggling to keep track of their token usage.
- The struggle to set prices for AI services has significant implications for the development and implementation of AI technology.
If companies can find a way to manage the costs of AI services, it could lead to a new era of innovation and growth in the industry. With the right pricing models in place, companies could focus on developing new and exciting AI technologies, rather than worrying about the cost of tokens.
If companies are unable to manage the costs of AI services, it could lead to a decline in the development and implementation of AI technology. This could have significant implications for the future of the industry, and could even lead to a backlash against the use of AI.
