Bark brings Ark to Bitcoin mainnet

Second has opened its Ark server, Bark SDK and early wallets for self-custodial off-chain bitcoin payments.

Second has announced that Bark is live on Bitcoin mainnet, alongside a public Ark server, a developer SDK and several apps users can try. The central fact is narrow and verifiable: this is not just another testnet demonstration. Second is making Bark, its implementation of the Ark protocol, available for real bitcoin payments while aiming to keep users in control of their coins. The company frames the launch as an attempt to reduce the friction of on-chain payments and avoid the channel management that still makes Lightning difficult for many non-specialist users.

Ark is a Bitcoin layer-two approach built around virtual UTXOs. A UTXO is an unspent transaction output, the basic unit a Bitcoin wallet later spends. A virtual UTXO represents a user’s claim through pre-signed off-chain transactions that can be brought back on-chain if needed. In normal operation, an Ark server coordinates payments and batches activity, while the client retains the keys. That is the design tension: the server does heavy coordination work, but the user should still have an emergency exit instead of simply trusting a custodian. Second also emphasizes interoperability with Lightning and regular on-chain payments, which matters if Ark is to become a payment tool rather than an isolated experiment.

The launch is notable because it turns Ark from a promising design into a more practical developer surface. The Bark SDK includes a Rust core with bindings for Kotlin, Swift, React Native, Flutter, Go, Python and the web through WebAssembly. For product teams, that packaging matters as much as the protocol details. A payment layer becomes useful only when wallet makers, merchant services and mobile apps can integrate it without rebuilding every low-level component. Second also points to Barkd, a daemon for more technical deployments, and to several wallets available at launch.

There are still obvious caveats. A mainnet launch does not prove adoption, long-term liquidity, fee behavior under stress, or the user experience of exits when something goes wrong. Ark also asks developers and users to understand a different security model from both direct on-chain payments and Lightning channels. Still, the signal is worth tracking: Bitcoin now has a mainnet Ark implementation that does not require a consensus change, but tries to improve everyday self-custodial payments. If Bark works in practice, the impact will not look like a new asset launch. It will look more like infrastructure: making small bitcoin payments easier to build, receive and spend without handing custody to an app operator.