Celestia

Celestia does not run smart contracts and does not process transactions. It orders data and guarantees that data was published — nothing else. That deliberate narrowing is the entire product, and it exists because data availability, not execution, is the expensive part of running a rollup.

Data availability sampling

Verifying that a large block was fully published normally means downloading all of it. Celestia lets a light node sample small random pieces and, after enough successful samples, conclude with high confidence that the whole block is available. Security improves as more light nodes sample, which inverts the usual relationship where more users mean more load.

Why separating execution matters

A monolithic chain must agree on what transactions mean, which requires every node to execute them. If a chain only guarantees that data exists and was ordered, execution can happen anywhere, and many chains can share one availability layer. Celestia is a bet that this separation is the correct architecture for the next generation of chains.

Sovereign rollups

A chain posting data to Celestia keeps its own rules and upgrade path rather than deriving them from a settlement contract on another chain. It gains published data guarantees without inheriting governance — sovereignty and data availability, decoupled.

What TIA pays for

Blockspace on Celestia is bought with TIA by the chains posting data, and TIA secures the network through staking. Demand is therefore tied to how much data other chains publish, which is an unusually direct relationship between token demand and network usage.

The open question

The modular thesis is architecturally sound and commercially unproven at scale: it depends on many chains choosing to outsource data availability rather than posting to Ethereum or handling it themselves. That competition is ongoing and the outcome is not settled.