Starknet Mainnet is having a hot break.
Starknet, an Ethereum layer-2 network that uses zero-knowledge (ZK) coils, is experiencing a new mainnet disruption as the project enters 2026.
In X's post, the Starknet team said the network was experiencing downtime and that engineers were “actively investigating the issue and working to restore full functionality as soon as possible,” without immediately specifying the cause. At the time of writing, the network experienced a delay of more than two hours.
Starknet is a ZK-rollup-based layer 2 that distributes off-chain transactions and encrypts cryptographic data to Ethereum, inheriting Ethereum's base-layer security for high-cost and low-fee smart contracts, decentralized finance, and gaming applications.
The project Bitcoin DeFi, or BTCFi, arc promoting itself as an infrastructure to bring Bitcoin-related financial applications to the Ethereum ecosystem. Despite the network outage, the price of the STRK token remained stable at the time of writing.
Related: Ethereum's first ZK-rollup ZKsync Lite will retire in 2026
It's not the first time Starknet's minnet has been taken down.
The incident took place in Following a series of outages in 2025, Starknet's reliability will be closely scrutinized. In September, a major update known as Grinta (v0.14.0) led to an extended mainnet outage that halted block production and necessitated two chain reorganizations that restored an hour of activity and forced users to resubmit affected transactions.
That scenario was linked to a series of issues following a multi-hour outage earlier in 2025, and external monitors recorded several instances of slow or stopped creation during the year.
A Starknet incident report on the September event said the downtime associated with Grinta lasted approximately nine hours and traced the problems to a series of issues, including failures in Ethereum RPC providers and bugs affecting serial behavior, prompting the team to make architectural changes and expand monitoring.
A Starknet representative told Cointelegraph that the team is working on fixing the new incident.
