Source
Why your AI bill tripled while token prices fell 75%
AI costs are increasing despite a 75% drop in token prices because enterprises are adopting agentic workflows that require significantly more tokens than simple chatbots.
Why it matters
Unpredictable AI spending and poor cost visibility can lead to project cancellations and reduced budgets if engineers cannot prove the return on investment for specific tasks.
The details
- Agentic coding can be 1,000 times more token-hungry than simple code-chat exchanges. - 81% of enterprises run three or more AI models. - Only 23% of AI projects launched last year met original ROI objectives.
What's next
Organizations should shift from tracking tokens to measuring cost per completed task using specific telemetry fields.
Show entities and relationshipsHide entities and relationships
In this article
Key connections
Stanford Digital Economy Lab is related to Agentic Coding
Research from the Stanford Digital Economy Lab puts agentic coding as much as 1,000 times more token-hungry than a simple code-chat exchange
Elastic is related to Observability
In a 2026 Elastic survey of enterprise IT organizations, 85% said they plan to implement large language model observability.
Get the weekly recap
The stories like this one, picked and explained — once a week, straight to your inbox.