ClearML Enterprise v3.26 brings powerful upgrades across model deployment, NIMs container deployment, and dataset management – all part of our end-to-end platform for managing and scaling AI in the enterprise.
Static Routes for Deployed Endpoints
This release introduces Static Routes – allowing you to define custom, persistent endpoints for deployed models. This feature acts as an abstraction layer between client-facing URLs and the actual models behind them.
With Static Routes, you can:
- Route multiple models behind a single, stable URL
- Swap or upgrade models without disrupting consumers
- Load balance across model replicas or versions
- Use a UI to track usage and manage all public routes
Static routes are particularly useful in production environments where models are frequently updated but workflows depend on consistent endpoints.
New Model Deployment Options
ClearML now supports the SGLang engine, joining existing support for vLLM and llama.cpp. This lets engineers deploying LLMs decide on the right engine for their use-case.
In addition, there’s a new UI for deploying NIM models, simplifying launch, monitoring, and management of NIM-based inference services.
General Improvements
- New admin panel with user activity insights (last login, credential creation, UI activity)
- Hyper-Dataset schema now visible in the UI
- Global search now supports quick filters
- Search added to the Queues table
- Dozens of UI enhancements and bug fixes
For full implementation details and benchmarking data, check out the full release notes.
Want to see it in action? Request a demo at clear.ml/demo