Attach storage to a sandbox and an agent's files and working state survive restarts, resumes, and redeploys. State is persistent by design, so an agent can remember what it did and pick up across sessions without you building a state layer around it.
Under the hood, every sandbox runs its root filesystem in memory for the fastest possible reads and writes. When a sandbox scales to zero, Blaxel snapshots its full filesystem and memory to warm storage, so it resumes from standby in about 25ms with exact state intact, even after weeks idle. When you need data to outlive the sandbox itself, you move from snapshots to durable volumes or a shared drive.
Three building blocks cover the full range, from instant in-session memory to long-term retention to real-time sharing across agents.
01/IN-SESSION MEMORY
An in-memory filesystem with snapshots kept until you delete them. Full filesystem and memory state is captured at scale-down and restored on resume, so a sandbox comes back exactly where it left off and running processes continue. Best for fast, in-session working memory.
02/LONG-TERM DURABLE
Durable mounts that retain data for years on a redundant backend. A volume is block storage attached to a sandbox at creation, and it survives sandbox destruction and recreation, so it is the cost-effective way to persist files from weeks to years. Sizes are set in MB and can be resized as you grow, and volume templates let you start from a pre-populated filesystem. Volumes are regional, so the sandbox or agent that mounts a volume runs in the same region. Best for long-term retention, caches, and model artifacts.
03/SHARED DISTRIBUTED
A distributed filesystem for real-time, multi-agent collaboration. A drive mounts into multiple sandboxes or agents at once, including while they are already running, with concurrent read-write access and built-in replication for durability. Unlike a volume, which attaches to a single sandbox, a drive behaves like a shared cloud filesystem in the file tree. Use it to pass files between sandboxes, share datasets across agents, store tool-call outputs and context histories, or cache common dependencies. Best for many agents working on the same files. Agent Drive is currently in private preview.
[DECISION MATRIX]
Fast in-session working memory
Long-term retention, caches, artifacts
Real-time collaboration on shared files
INFRASTRUCTURE LIMITS
Storage scales with the same tier-based quota system as the rest of the platform, up to 512 TB of storage and 100,000+ concurrent sandboxes or agents. Tiers upgrade automatically as you top up credits.
No base subscription. Storage is fully usage-based, and you start with up to $200 in free credits, no credit card required.
$0.20/ GB-MONTH
Memory and filesystem snapshots that keep suspended sandboxes ready for about 25ms resume.
Usage-based
Durable, long-term storage billed per GB-month, scaling up to 512 TB under the tier-based quota system. See the pricing page for the current rate.
$0.045/ GB-MONTH
Templated deployment images stored for fast, repeatable environment creation.
Private Preview
Pricing and storage quotas are being finalized; request access to try it.
You only pay for what you store. Suspended sandboxes stop billing compute and are charged at the snapshot storage rate to stay resume-ready. All sizes are binary gigabytes (GiB), where 1 GB = 2^30 bytes. See full pricing at https://blaxel.ai/pricing.
