The OpenCost project, an open-source cost and resource management tool hosted by the Cloud Native Computing Foundation (CNCF), has published a year-in-review reflecting on its progress in 2025 and outlining priorities for 2026. The update highlights an active release cadence, expanded capabilities including an AI-ready MCP server, community growth through mentorship and contributions, and plans to extend the project's data model and cost tracking features.
In 2025, the OpenCost community delivered 11 releases that enhanced usability and expanded functionality. Notable additions included the ability to run OpenCost without requiring Prometheus by using environment-variable configuration or the beta Collector Datasource, a generic export framework for cost data, and a diagnostics system with health-tracking and export capabilities. OpenCost also improved multi-cloud cost tracking, with contributions from providers such as Oracle and DigitalOcean helping extend support for tracking cloud and multi-cloud metrics. These releases aimed to make cost transparency more actionable across Kubernetes environments, benefitting both users and contributors of the project.
A significant milestone in 2025 was the introduction of the OpenCost MCP server, enabling AI agents to query cost data in real-time using natural language. This integration enables the automatic analysis of spending patterns across namespaces, pods, and nodes, allowing teams to generate cost reports and recommendations without manual queries. The MCP server emerged as a default component that outputs clear, step-by-step suggestions for cost optimization, blending cloud cost management with emerging AI ecosystems to meet the needs of more automated FinOps workflows.
Community activity also featured prominently. The OpenCost project continued its mentorship efforts through the Linux Foundation's LFX program, with mentees contributing integration tests for enterprise readiness and advancing OpenCost's Data Model 2.0 (KubeModel), a foundation for scalable, accurate cost tracking across dynamic Kubernetes resources. Contributors also focused on documentation, UX improvements, and broader community engagement, reinforcing OpenCost's developer-friendly orientation and ecosystem growth.
Looking ahead to 2026, the project is planning enhancements to support AI usage cost tracking as machine-learning workloads increasingly factor into cloud spending. Improving supply chain security around cost data is also a priority, along with iterative refinement of the KubeModel framework to better reflect the complexity of Kubernetes resource behavior. Participation in upcoming KubeCon + CloudNativeCon events remains a key part of the project's strategy to grow awareness and adoption among cloud-native practitioners.
OpenCost's role as a CNCF incubating project underscores its relevance in a cloud-native landscape where cost visibility and resource allocation have become central to operational and financial governance. By standardizing Kubernetes cost reporting and integrating with AI-driven tooling, OpenCost says it aims to support both FinOps and engineering teams as they navigate increasingly complex multi-cloud workloads in 2026 and beyond.