Qwen family
A candidate for multilingual and general-purpose workloads, subject to model-specific validation.
MODELS / OPEN WEIGHTS
Your model. Your dedicated runtime.
We deploy a model selected for your use case and verify its hardware fit before provisioning. Availability depends on licensing, model access, and the required GPU capacity.
Llama family
A candidate open-weight family for enterprise deployments, with license review before use.
DeepSeek family
A candidate for reasoning-oriented workloads where hardware and model terms fit.
These are candidate model families, not a promise that every variant fits every GPU or license scenario.