Vultr expands serverless inference with retrieval tools

A serving platform with retrieval and application integration

Vultr's October 17, 2024 announcement described automatic resource scaling, private GPU deployments and retrieval over customer data. The expansion added a managed route from stored documents to model responses, with an API compatible with existing clients. Contemporary coverage reported the same service expansion. [1][3]

This followed a March 2024 private-reservation beta. October's event therefore records the expanded inference service, rather than an unsupported first-ever launch date or a commercial relationship implied only by API compatibility. [2][1]