Nearly 16% of tracked enterprise software companies have adopted usage-based AI pricing as of late 2025, more than doubling from 2024. Major players including Adobe, Salesforce, and HubSpot are abandoning traditional per-user subscription models.
The shift reflects a fundamental change in how software vendors monetize artificial intelligence capabilities. Traditional seat-based pricing, where companies charge per user regardless of AI usage, no longer fits the economics of AI-powered features that consume variable computational resources.
Out of 500 tracked software companies, 79 have implemented usage-based AI fees—a dramatic increase from roughly 30-40 companies in 2024. The change signals industry-wide recognition that flat-rate models cannot sustain the infrastructure costs associated with AI workloads.
HubSpot, Adobe, and Salesforce represent the vanguard of this transition. These enterprise stalwarts serve millions of customers and process massive amounts of data through AI systems daily. Their adoption of usage-based pricing creates pressure on competitors to follow suit or risk losing pricing flexibility.
Usage-based models typically charge customers for AI features consumed—such as the number of AI-generated emails, automated analyses, or API calls processed. This approach aligns costs with actual consumption and creates direct incentives for customers to optimize their usage.
The shift presents both opportunities and risks. Vendors gain better cost alignment and potential upsell opportunities. Customers, however, face unpredictable monthly bills if usage spikes unexpectedly.
Analysts note the trend will likely accelerate as more companies invest heavily in AI infrastructure. Companies that fail to shift pricing models may face margin pressure or competitive disadvantage. The remaining 421 tracked companies not yet implementing usage-based AI fees will face increasing pressure to adapt.
This represents one of the most significant pricing model transitions in enterprise software since the shift to cloud-based subscriptions two decades ago.
Z.ai released GLM-5.3's weights on Hugging Face under a new license that requires large companies to undergo security review before hosting the model. The change marks a departure from the standard MIT license.
Anthropic has introduced the Model Hardware Standard (MHS), a unified interface enabling AI agents to operate robotic arms, lab instruments, and other physical devices. Early testing shows integration time has dropped from weeks to hours.
Open-weight AI companies—those releasing freely available models—are attracting major acquisition interest from tech giants. The trend reflects growing capital investment in the business model of distributing AI models at no cost.
Google Deepmind has upgraded its Co-Scientist AI system to autonomously plan experiments, operate lab equipment, and publish scientific papers. The Gemini-based multi-agent platform demonstrated experimentally validated results across materials science, chemistry, and medical AI development.