A few weeks ago, we announced the Ohm v18 beta, which involved a complete rewrite of the core parsing engine. Since then, we've implemented even more performance improvements: v18 is now more than 50x faster for real-world grammars while using about 10% of the memory.