The sixth FinOps X (if you count the one in Barcelona) took place this week in sunny San Diego, CA. This year’s event was all about AI spend. Rapid adoption of generative AI SaaS (e.g. Claude, CoPilot), AI cloud infrastructure (e.g. GPUs), and AI cloud services (e.g. Amazon Bedrock) have skyrocketed concerns about managing their unpredictable, token-based costs.
The conversation – and concern – about AI spend has ballooned so quickly that it is now a boardroom discussion. Top-down mandates to reign in AI costs are putting pressure on FinOps practitioners to quickly learn a whole new set of capabilities – including model routing optimization, cache optimization, and context window optimization.
FinOps Foundation leaders have established a new Tokenomics Foundation to focus specifically on bringing clarity and control to AI token costs. Flexera is proud to be a founding member of the new foundation so we can partner with the community to develop common language, best practices, and to inform our roadmap so we can better serve our customers’ changing needs.
Flexera announces full stack AI cost management
At FinOps X, Flexera made two exciting announcements around support for AI cost management. See the keynote recording.
Flexera is building the only AI cost management platform spanning agents, models, data platforms, and compute. The first milestone is complete visibility of all AI spend so you can understand how much is being spent on what.
“We need a new AI operating model where we understand the complete economics of the AI stack, and we enable AI optimization to work in our favor,” said Becky Trevino, Chief Product Officer. “Flexera enables this new operating model by giving our users the ability to measure and benchmark usage and optimize consumption costs to affordably run AI at scale,” she added.
Becky Trevino, Chief Product Officer, shares Flexera’s new capabilities for AI cost management.
Becky also shared that FinOps Assist is being rolled out in Flexera One Cloud Cost Optimization. FinOps Assist is an ‘AI for FinOps’ feature that answers questions and generates reports from natural language prompts, enabling non-technical users to gather actionable insights relevant to their role. FinOps Assist will evolve into a multi-agent orchestrator that routes tasks across specialized agents.
Flexera announces FinOps Assist, an AI-powered assistant in the Flexera One console.
- See the press release with Flexera’s AI-related announcements
A new pricing model – pay for outcomes with ProsperOps+
Meanwhile, many organizations are just beginning their cloud and FinOps journey. They are discovering a robust industry with a sea of tooling vendors eager to show them positive return on investment.
ProsperOps and Flexera have decided to make your FinOps tooling decision very easy. We’ve built a positive ROI into our pricing model when you buy ProsperOps+, a bundle offering of our industry-leading solution for managing commitment-based discounts (ProsperOps) together with any of Flexera’s FinOps solutions, including Ocean, Elastigroup, CCO, DCO, or CLM.
With ProsperOps+, you only pay for quantified savings outcomes. We collect a small portion of either the savings generated through autonomous commitment management, or of the costs you avoided paying by using our autonomous usage optimization tools.
Customers with ProsperOps+ actually add money back to their budgets that can be reinvested in new initiatives.
- Read more about our outcome-based pricing offer, ProsperOps+
Erik Carlin, Co-founder and Chief Product Officer at ProsperOps, announces a new pricing model where ProsperOps + Flexera customers only pay for quantified savings outcomes.
Unifying rate and workload optimization for the first time
Last, but certainly not least, Erik Carlin announced something the FinOps industry has never seen before. Our exciting new feature brings autonomous rate and workload optimization together to deliver maximal savings. Rather than competing, as can happen when rate and workload automations are running in parallel, Unified Autonomous Optimization brings the two together under a ‘single brain’ that optimizes both workload usage and commitment discount usage (i.e. rate optimization) in tandem.
When customers buy a ProsperOps+ bundle of either ProsperOps + Ocean, or ProsperOps + Elastigroup, they’ll get fully autonomous rate + workload optimization that delivers maximal savings without disrupting VM-based or container-based workloads.
Erik Carlin unveils unified rate and workload optimization, a capability that no FinOps vendor has offered before.
Unified Autonomous Optimization is generally available today. New customers can sign up here.
- Read more about Unified Autonomous Optimization
Two customers have already achieved a 1+1=3 savings effect with Unified Autonomous Optimization. Duolingo applied this capability to their largest Kubernetes cluster and saw a 5% increase in their Effective Savings Rate (ESR), which, on $1M in monthly usage translates to an incremental $600,000 of savings a year to be reinvested in innovation.
Tealium applied Unified Autonomous Optimization to their clusters at scale, and as a result, for every $1 of on-demand cloud spend, they are only paying 37 cents.
Duolingo and Tealium are achieving a 1+1=3 savings effect with unified rate + workload optimization
Flexera has been a leader in FinOps for years, but recent acquisitions greatly expand our capabilities. We are excited to see the first proof points that bringing these separate solutions together can help drive higher savings outcomes for our customers.
Closing
As FinOps evolves to meet the demands of AI and increasingly complex cost structures, one thing is clear: fragmented approaches are no longer enough. From full-stack AI cost management to Unified Autonomous Optimization, Flexera and ProsperOps are helping organizations meet today’s complexity with an integrated set of intelligent tools. We’re just getting started, but if FinOps X 2026 proved anything, it’s that the future of FinOps will be defined by automated optimization, and the ability to turn technology costs into quantified value.
