solx is a new LLVM-based Solidity compiler for Ethereum. Today it eliminates the long-standing stack-too-deep error without introducing semantic changes or a separate pipeline, so contracts behave exactly as they do with solc’s default path—just without the failure and at lower runtime gas cost.
It already passes the full solc internal test-suite plus our own corpus of 2 000 + contracts drawn from Uniswap V4, Aave, Solady, and other production projects. solx is built on the same LLVM fork that has powered ZKsync in production since 2021.
As for performance, ≈ 92 % of benchmarks show lower gas usage than solc (per project breakdown), while the emitted bytecode is only ~6 % larger on average (unaggregated data). Given these results, we label solx beta: suitable for non-mission-critical deployments today, with more features, security hardening, and size tuning on the way before a stable release.
We invite you to try solx at https://solx.zksync.io. If you’d like deeper context, see our three Mirror posts:
- solx Beta – Status Update – current capabilities, numbers, roadmap.
- solx Alpha – Early Walk-through – published with our first release in May, though status is outdate, the text provide code examples where solx is better than solc.
- Why LLVM Matters for Solidity – how LLVM let us fix stack-too-deep in two months and keep our fork to ~9 k LoC.
Please share your feedback—positive or constructive.
Does Ethereum need a second Solidity compiler? What does solx still lack for you to adopt it in production?