I did some recent testing that showed both Lua and LuaJIT-joff (its interpreter-only mode) to be about 2x faster than Python. Both PyPy and full-on LuaJIT were about 10x faster.

Years ago, I even found Ruby to be faster than Python. This was back in the Ruby 2.0 / Python 3.5 days - I'd be interested to know if it's still the case.