Fresh From the Labs: CloudHealth Product Updates September 2026
About the Author: Cathal Cleary - Product Leader, CloudHealth | Governing Board, FinOps Foundation
Mastering AI Spend, MCP Server Release
It’s October, the air is getting a little cooler and the leaves are changing in the Northern Hemisphere, if you’re in the Southern Hemisphere you’re looking forward to Summer! The buzz around generative AI continues to dominate conversations—the CloudHealth team attended the first Tokenomicon in Amsterdam last week. The Tokenomics Foundation launched the inaugural State of Tokenomics survey, which revealed many things some expected and some unexpected. Amongst the expected, we can see that the majority of organizations struggle to justify AI spend to the CFO. For the minority that can justify AI spend, the pattern shows that they are able to attribute spend across workloads and teams. This month’s batch of updates from CloudHealth should greatly help with this visibility. But we’re far from done. Watch this space for more exciting updates and deeper visibility from the CloudHealth team in the coming months.
MCP brings the power of CloudHealth’s rich datasets to your favourite AI harness. Reduce the time it takes to prepare your monthly updates from hours to minutes, prepare that advanced analytics report on spend trends with a few AI prompts. Get creative with your prompt and see what you can do with CloudHealth data and MCP.
Alongside AI enhancements, we are also rolling out advanced margin sharing controls for our Managed Service Provider partners, ensuring your multi-cloud operations remain efficient, profitable, and fully optimized.
Read on for this month's product updates designed to elevate your FinOps practice.
Enterprise AI Visibility & Integration
Combined Multi-Provider AI Dataset
What's New: CloudHealth now supports the "Multi-Provider AI Cost and Usage" dataset. This addition takes cost and usage data from all datasets containing user-level attribution—currently supporting OpenAI and Anthropic—and normalizes them into one single AI schema. We will continue to expand support for more model providers, including Copilot and Bedrock, very soon. To get you started instantly, three standard reports are available out of the box: Multi-Provider AI by Organization, Model, and User.
Why It Matters: As enterprise AI adoption scales, tracking individual usage across different vendors quickly becomes an administrative burden. By unifying diverse AI provider data into a single, clean schema, FinOps teams can seamlessly execute chargebacks, spot runaway costs at the individual user level, and accurately measure the cost of their AI initiatives without exporting and merging disparate spreadsheets.
User-Level Data in Tokenomics Dashboard
What's New: The existing Tokenomics dashboard has been upgraded to natively include cost and usage data from OpenAI and Anthropic, pulling directly from the new Multi-Provider AI Cost and Usage dataset. A dedicated dashboard section called "Frontier Lab AI" has been added to house these specific widgets. For customers who have not yet configured their Data Connects for these models, this section can be easily collapsed to keep your dashboard clean.
Why It Matters: Visualizing your AI footprint should be immediate and effortless. This update puts critical AI consumption trends front and center, providing executive-friendly visualizations that make it easy to digest user-level spending patterns and optimize token usage at a glance.
CloudHealth MCP Server
What's New: CloudHealth now integrates directly with leading AI clients like Cursor, GitHub Copilot, and Claude via the Model Context Protocol (MCP). With just a few lines of JSON, users can connect the CloudHealth MCP to any agent and model to execute complex reporting tasks. To support the MCP server, CloudHealth has introduced the concept of Service Accounts to the platform for the first time. Currently they are only applicable to the MCP server, but we will expand their usefulness to APIs in a coming release.
Why It Matters: This integration represents a massive leap forward for operational efficiency. Instead of manually digging through data grids to build presentations, FinOps teams can use natural language prompts to perform heavy lifting. You can instruct your AI assistant to generate monthly review slides, create chart representations of growth trends and variances, execute chargebacks, or perform anomaly detection using your live CloudHealth data. It transforms raw financial data into actionable business artifacts in seconds.
Platform & MSP Enhancements
RI and Savings Plan Sharing for AWS Partner Billing
What's New: Partners can now set their margin as a percentage of the Reserved Instance (RI) and Savings Plan savings, rather than simply marking up the raw cost. A new RiSpSavingsShare rule in Custom Price Books allows partners to keep a defined percentage of the net savings generated from each RI and Savings Plan, seamlessly passing the remainder on to the end customer. Percentages can be uniquely scoped by service or by Savings Plan type, and the share is applied line by line on the RI fee and Savings Plan recurring fee lines.
Why It Matters: Managed Service Providers need flexible, value-based billing constructs to demonstrate their impact and maintain healthy margins. This highly requested feature allows partners to directly monetize their cloud optimization efforts. By sharing the savings, you create a perfect win-win scenario: customers receive lower bills through expert commitment management, and partners are fairly compensated for their strategic FinOps oversight.
Jumpstart Your October
We are thrilled to bring these advanced AI tracking and partner billing capabilities to your FinOps practice this month. If you aren't leveraging these new capabilities yet, log into the platform today to jumpstart your cloud optimization journey.
If you would like to get more information about CloudHealth, start a free trial, or have questions for our team of FinOps experts, schedule some time with us!
Until next month—happy building!