Bitcoin Silent Payments Require Only 8 MB Daily to Scan

A new measurement of Bitcoin's Silent Payments scan data shows the bandwidth side of mobile private receiving is more manageable than feared. The BlindBit Oracle v2 project tracked every block from height 709,656 through 965,089, an inclusive span of 255,434 blocks from shortly after Taproot activation through September 2026. Across that range the full scanning payload totaled 15.08 GB, averaging 59 KB per block, while a wallet following the chain near the dataset's upper end consumes about 8 MB per day. That puts the download cost roughly in the range of a short streaming video.

Why it matters

Silent Payments (BIP-352) let a user receive BTC at a reusable static address while keeping outside observers from linking payments to that address on-chain. A sender derives a fresh Taproot destination from the recipient's public data, and the recipient's wallet later scans eligible transactions to find its output. Doing that scan on a phone, without running a full node, has been the unresolved part of the design.

Market impact

The numbers take the most-cited objection off the table. The 8 MB-per-day figure sits well inside any modern mobile data plan, and a modeled Taproot-only filter route lands at about 7.14 GB across the same history. But the same measurement exposes the next layer of risk: a light wallet depends on an indexer to deliver complete scanning data, and a single omitted tweak can leave a real payment invisible without any user-visible error. Cake Wallet's July 2025 issue traced how a client can save a later scan height after receiving incomplete data and treat earlier history as finished, so switching to a different honest server later does not automatically trigger a rescan. The project's chained-commitment proposal would let two servers be cross-checked for completeness, with externally anchored checkpoints exposing equivocation. Wallet adoption remains fragmentary: Sparrow 2.5.0 added Silent Payments with default Frigate server selection in May 2026, BIP-352 itself is marked Complete for the core protocol, and the converged light-client spec sits at pre-draft v0.1.