“Model inference”

Real-time AI infrastructure for deploying voice agents, video models, LLMs and other AI workloads, with autoscaling.

Model router that predicts which LLM to call for each request, aimed at cutting coding-agent costs.