A LinkedIn post from Mavvrik discusses observations from Google Cloud Next ’26, emphasizing the rapid growth in large language model usage and related infrastructure needs. The post cites Google figures such as 330 customers each processing more than a trillion tokens over the past year and models handling 16 billion tokens per minute via direct API, up from 10 billion last quarter.
Claim 55% Off TipRanks
- Unlock hedge fund-level data and powerful investing tools for smarter, sharper decisions
- Discover top-performing stock ideas and upgrade to a portfolio of market leaders with Smart Investor Picks
The post suggests that this expansion is driving substantial investment in TPUs, GPUs, data platforms, agents, and APIs, with cost implications at every layer of the stack. It highlights Google’s new Gemini Enterprise Agent Platform, particularly an “optimize” pillar featuring Agent Observability with standardized, OTel-compliant telemetry across agents, tools, and API handoffs.
According to the post, observability is only one dimension of managing agents at scale, and understanding cost remains a separate challenge. Mavvrik positions its offering as addressing cost attribution, chargeback, and accountability across cloud, GPUs, inference, agents, SaaS, and on-prem environments, and notes that it is part of the Google Cloud ecosystem and available via the Marketplace.
For investors, the commentary points to rising enterprise spend on AI infrastructure and an emerging need for granular cost management as usage scales. Mavvrik’s alignment with Google Cloud and focus on AI cost governance may support demand for its platform if token volumes and agent-based workloads continue to grow, potentially enhancing its strategic relevance within the broader cloud and AI tooling ecosystem.