Soda Labs develops open-source Garbled Circuit-based SMPC (Secure Multiparty Computation) tools that enable confidential smart contracts and encrypted data processing directly on EVM-compatible networks. The Garbled Circuit–based confidential compute engine operates in two modes: integrated and plug-and-play. In the integrated mode, the MPC protocol is embedded within the consensus layer to produce a fully fledged, privacy-preserving L1 or L2. The plug-and-play mode, on the other hand, is run by an external network (called Bubble) to enable privacy on existing chains with minimal integration effort, while also supporting cross-chain functionality.
Soda Labs – MPC SDK & Confidential Smart Contracts (Garbled Circuit for EVM)
Shielding - Garbled Circuit-based implementation achieves shielding capabilities similar to those of ZK
Not a substitute for
ZK-based L2 privacy (e.g., Aztec, Scroll)
MPC or TEE for custody
Architecture
Soda Labs provides libraries for private instruction set analogous to the set supported by the EVM.
In runtime, once a private instruction is encountered (e.g., private-ADD256) it triggers the execution of a garbled circuit for that specific instruction (the circuit receives encrypted inputs and produces encrypted output).
The Evaluators (i.e., the parties that are responsible for the confidential computation) are assumed to have in their posession garbled circuits for all types of instructions, and therefore they take part in a continueous process to produce those garbled circuits and maintain an inventory.
Given that inventory of garbled circuits, processing a block that demands privacy-preserving workload is done non-interactively, following a constant number of rounds for soldering the relevant garbled circuits.
Privacy domains
Selective disclosure for regulatory compliance and audit trails
Enterprise treasury operations with confidential payment flows
Standard AES encryption at the contract and variable level
Supports hybrid models of Garbled Cirtcuit + ZK for public auditability
Enterprise demand and use cases
Financial institutions seeking on-chain confidentiality with deterministic settlement.
Private vaults, confidential lending, or yield strategies.
Technical details
EVM-compatible GC runtime (gcEVM)
SDKs for Solidity and python
Strengths
Native EVM integration
Strong cryptographic research pedigree
Relies on standard and time-tested & PQ-secure encryption scheme: AES
GC provides the lowest latency for general-purpose confidential computation
Cheap: runs on low-end machines (no GPUs/FPGAs/ASICs are required)
Risks and open questions
Interoperability between GC networks still emerging
Standardized and audited ERC contracts is in the work
Secure under the assumption of threshold number of honest participants (just like any other MPC/FHE)
