Robin Linus Publishes the BitVM3 Garbled-Circuit Design
Robin Linus published a BitVM3 design based on garbled circuits, describing a path to make SNARK verification on Bitcoin dramatically more efficient than BitVM2. The work remained an experimental protocol design rather than a production-ready bridge system.
On June 15, 2025, Robin Linus published a new BitVM design built around garbled circuits, calling the approach BitVM3. BitVM research seeks ways to verify complex off-chain computation with Bitcoin transactions without changing Bitcoin’s base-layer consensus rules. The new proposal focused on one of the largest practical obstacles in earlier BitVM designs: the cost and complexity of proving that an off-chain computation was performed correctly when a dispute reached the Bitcoin blockchain.
Linus wrote that the BitVM Alliance was pivoting toward garbled circuits because they could make SNARK verification on Bitcoin more than 1,000 times more efficient than BitVM2. BitVM3 was presented as one point in a broader design space, balancing implementation complexity, communication overhead and proving cost. The design also attempted to solve a weakness in an earlier garbled-circuit proposal by allowing circuits to be verified in plaintext and then reblinded, avoiding very large disprove scripts or expensive correctness proofs.
BitVM3 was research, not deployment. It did not create a new Bitcoin opcode, activate a soft fork or launch a production bridge. Its importance was as a technical step in the continuing effort to build more expressive verification systems around Bitcoin while preserving the base protocol. The proposal also illustrated how rapidly the BitVM field was evolving after its original introduction, with bridge security, trust assumptions and on-chain dispute costs becoming active areas of Bitcoin engineering research.