Tapedrive runs its own network of storage nodes, much like a blockchain runs its own validators. Instead of validating transactions, the nodes store files, and they keep proving your data is still there.
To you, it just looks like S3: same tools, new endpoint.
Note
Tapedrive is in early access and invite-only.
Sign up for access, or join the Discord to follow development, and read the docs for the full picture.
Your data is distributed to many nodes on the network. Every large write is erasure-coded, onto a spool group. The network runs many groups, with no cap on the number of groups or nodes.
Storage overhead is about 2.8x the stored size, far less than other storage networks. Repair is bandwidth-optimal: a lost piece is rebuilt from small fragments read across many nodes, not from whole copies.
The S3-gateway means most existing tooling works after changing an endpoint and a credential.
When you want signed writes and verified reads straight from your own code, the CLI and SDKs speak the network natively.
Install the CLI. Prefer code? The SDK quickstart walks the same flow in six languages.
Already on S3? Point your existing tooling at the S3-gateway and keep your workflow.
Storage nodes stake TAPE, hold slices of the network's data, and earn for proving they still have them, on hardware you own or rent. The node setup guide takes you from a bare Linux machine to a registered, earning node.
- Docs: from the quickstart to the white paper.
- When to use Tapedrive, and when not to.
- Explorer: watch the network live.
- Discord and X.
Looking for the earlier Proof-of-Work version? It lives on the
PoW branch.