Skip to main content
Span batches are a batch format, introduced in the Delta network upgrade, that encodes a span of consecutive L2 blocks in a single batch. They reduce the overhead of posting L2 data to L1, which lowers data availability costs — especially for sparse, low-throughput chains where fixed per-block overhead dominates the submitted data.

​
Why span batches exist

Before Delta, the only batch format was the singular batch: one batch per L2 block. Every block — including empty blocks on a quiet chain — carried its own batch overhead to L1. A span batch instead represents a whole span of consecutive L2 blocks in a more efficient encoding, while preserving the same consistency checks as singular batch data. That makes the marginal cost of including an additional (especially empty) block in the submitted data small. The full format is defined in the span batches specification.

​
How span batches are used

Span batches are opt-in for the chain operator: the op-batcher selects the batch format with its --batch-type flag (0 for singular batches, the default; 1 for span batches). On the verification side, the op-node derivation pipeline accepts span batches only once Delta is active on the chain — it drops any span batch whose L1 origin predates Delta activation. Singular batches remain valid after Delta, so a chain can switch between the two formats at any time. To turn span batches on for your chain, follow the how-to guide: Enable span batches.