I doubt it unfortunately - the FTS operators are too heavyweight to be inlined (or benefit from inlining), so JIT isn't likely to help a ton. Parallelism can h

I doubt it unfortunately - the FTS operators are too heavyweight to be inlined (or benefit from inlining), so JIT isn't likely to help a ton. Parallelism can help however, depending on what your bottleneck is (if the query produces a bitmap indexscan the heap access part can be parallelized).