Triton Inference Server
Serve models from multiple frameworks on cloud, edge, GPU or CPU systems.
What’s included for free?
Triton Inference Server is ongoing-free under BSD-3-Clause with no vendor request quota; operators supply hardware, models and deployment infrastructure.
Where it runs: Self-hosted
Provider information ↗Checked 2026-11-01
Before you use it
BSD-3-Clause covers the server; container images, NVIDIA components, model backends and weights include separate notices.
Provider terms can change. Check current pricing and usage rights before starting a project.