A Note to Our Users, Customers and Partners: Welcome to Klique :)

By Moses Guttmann, Co-founder and CEO, and the ClearML team

ClearML Enterprise is now Klique, hurrah!

A few years ago, we spotted a fundamental gap in orchestration tooling for data scientists. We built ClearML, which quickly became synonymous with MLOps. As industry requirements expanded toward broader AI management, ClearML evolved alongside them. We are now adopting a new name to reflect this market evolution.

Working with more than 2,100 organizations along the way, we recognize the demand for a silicon-agnostic, environment-agnostic AI platform that provides users flexibility and performance. Our customers are investing more than ever in people and infrastructure and automating with agents. As enterprises move from training models to running AI inference and autonomous agents company-wide, the need for effective resource optimization and full observability becomes critical.

Klique is that next move. It is the next generation of AI resource management built on the ClearML software foundation. It’s the same great team, expanding ClearML’s existing capabilities to address the challenges AI will present to organizations, by adding AI tools such as an LLM gateway and model routing to provide holistic cost control. As R&D is now everywhere with agentic tools, we are providing a solution for the entire expanded AI enterprise workforce, from technical consumers to AI builders and IT operators, on one platform to work and share together. This is an evolution. Everything you’ve used, deployed, and supported as ClearML is still there, now with additional agentic capabilities built on top. We’ve simply given it a name that matches where our vision and platform are headed.

What is Klique?

Klique is the control plane for enterprise AI, it manages GPUs and compute across hybrid, on-prem, and air-gapped environments on any silicon, and it controls which AI tools, models, and agents can use those resources, and at what cost.

On top of that foundation, we added what companies need to run AI and agents everywhere. The first piece is an LLM gateway: a single entry point for all model traffic in the company. The second is model routing, which sends each request to the best model for the job, so you have one place to control cost.

Any AI tool you already use connects directly to your internal resources, with budget controls and role-based access control (RBAC) tied to your identity provider. Any model, whether you built it in-house, took it from open source, or pay for it as a foundation model, is available from one platform with the same security and budget controls.

Klique is a platform where the people building AI, the people using it, and the IT teams running the infrastructure all work and share under the same rules.

Why this matters for enterprise teams

Once a company moves from training a few models to running inference and agents everywhere, resource optimization and observability stop being nice-to-haves. People want to open their favorite AI tool and just work, while IT needs to know what’s running, on which hardware, with which data, and at what cost. If you can’t see it, you can’t optimize it, and you’re ending up paying for idle GPUs and tokens no one is tracking.

ClearML open source and existing deployments stay the same

Nothing changes for ClearML open source or for existing deployments. This is an evolution, and everything you’ve deployed as ClearML is still there.

ClearML open source isn’t going anywhere. It stays free and supported on the ClearML GitHub repository. It covers the same lifecycle it always has: data management, experiment tracking, pipelines, scheduling, and orchestration for running jobs on any CPU or GPU.

Moving your ClearML Enterprise deployment to Klique

To upgrade your existing ClearML Enterprise deployment to a Klique deployment, contact your Customer Success Manager. They will walk you through the right path for your setup. Upgrading gives you Klique’s new expanded capabilities and keeps you current on feature updates, bug fixes, and security patches.

Thank you

We want to thank our open source community, our customers, and our partners. Every time you asked us for the next thing, you pushed us here. We built ClearML because the path from a data scientist’s machine to production should just work. Klique takes that same idea to the whole company.

Welcome to Klique. You can learn more at klique.ai.

FAQ

Is ClearML Enterprise being renamed? No. Klique is an expanded version of ClearML Enterprise, bringing new capabilities and expanding AI resources optimization beyond compute and into AI services.

Is ClearML open source going away? No. ClearML open source stays free and supported on GitHub, covering data management, experiment tracking, pipelines, scheduling, and orchestration.

Do existing ClearML deployments stop working? No. Current deployments keep running as they are. To get Klique’s new features, bug fixes, and security patches, you upgrade to Klique with help from your Customer Success Manager.

Where can Klique run? Klique runs in hybrid, on-prem, and air-gapped environments on any silicon, and it works with any in-house, open source, or foundation model.