Vultr added an Archival Object Storage tier to its cloud storage lineup, priced at $6 per terabyte per month with no retrieval fees and no minimum data retention period. The tier targets data accessed less than once a month: AI training datasets, disaster recovery backups, content archives, and application logs. Unlike most archival offerings from major providers, it makes data instantly accessible rather than subject to restore delays, according to Vultr.
The retrieval-fee structure is the pointed differentiator. Competing archival tiers from AWS Glacier, Google Cloud Archive, and Azure Archive all charge to retrieve data, sometimes at rates that make large-scale recovery operations significantly more expensive than the storage cost itself. Vultr's approach trades the cost-on-read model for a flat monthly rate, with egress bandwidth charged at $0.01 per GB globally after a per-subscription allotment. That formula gives teams operating large, infrequently touched datasets a more predictable bill, especially for disaster recovery scenarios where retrieval needs arise unpredictably.
Archival buckets can be added to an existing Vultr Object Storage Standard subscription or provisioned as a standalone subscription. Data already in the Standard tier can be set to migrate automatically into the Archival tier. The main architectural constraint is that the Archival tier does not support direct ingestion from outside Vultr Object Storage: standalone Archival deployments include an Unarchived bucket for uploads, with data marked to migrate into the Archival layer. When retrieved, data moves back to the Unarchived or Standard bucket it came from and sits there for seven days before automatically returning to the Archival tier.
Vultr claims 99.999999% durability and 99.99% availability, with data encrypted in transit and at rest. The offering competes primarily on price and fee transparency against a segment of the market where hidden retrieval costs have become a common friction point for teams budgeting long-term data retention at scale.
