Elevates cheq from a precise solver into a performance-tuned, configurable engine for large-scale charge equilibration. This release keeps the rigorous Rappé–Goddard core intact while introducing lossless speedups (cached geometry, lean memory reuse, streamlined matrix assembly) and an opt-in “fast hydrogen” one-pass mode that mirrors Open Babel’s behavior for H-rich systems when throughput matters more than exact nonlinearity. Parallel-friendly defaults embrace Rayon and faer’s global pool for effortless multicore gains. Updated options and documentation make these capabilities discoverable without burdening existing users, ensuring cheq stays both elegant and decisive for production pipelines and high-throughput simulation setups.