A model-serving platform for deploying custom models as production API endpoints.
It packages models with its own framework, Truss, and manages autoscaling and API publishing end to end, making it well suited for custom-architecture models beyond standard LLM inference.
© 2026 ITBGM