OrviScript 0.6.2 · verified locally
One million iterations. Near Node.js time.
OrviScript reached 115.9 ms end to end, compared with 102.3 ms for Node.js and 384.8 ms for Python on the same Ryzen 9 9950X3D machine.
Same program · one million iterations
Close to Node. Clearly ahead of Python.
Node.js 24.22.1± 24.5 ms · 84.1–218.2 ms
102.3 msOrviScript 0.6.2± 5.2 ms · 109.7–128.1 ms
115.9 msPython 3.15.2± 69.1 ms · 292.5–586.1 ms
384.8 msNode is about 1.13× faster than OrviScript.
OrviScript is about 3.32× faster than Python in this workload. Node measured 3.76× faster than Python.
Hyperfine detected statistical outliers, especially in the Node and Python results. This is a local comparative measurement, not a promise of identical performance in every program.
Inside the runtime
Less lookup. Almost no allocation.
The primary cost was not parsing. It came from boxed futures, allocation, and variable lookup by name. The fast path removes those costs for supported synchronous code.
What changed in 0.6.1
A fast path without splitting the language.
A program uses the bytecode fast path only when the supported subset preserves the same typing and execution rules. Other code automatically uses the compatible interpreter.
Bytecode fast path for synchronous programs
Indexed variable slots instead of name lookups
Fast if, while, blocks, operators, and print
Typed slots plus short-circuit && and ||
Automatic fallback for functions, collections, modules, and async
Method
One workload. The same result.
Hyperfine measurement from October 27, 2026 on Windows with a Ryzen 9 9950X3D processor: 5 warmups and 30 runs with the shell disabled. OrviScript 0.6.2, Node.js 24.22.1, and Python 3.15.2 each started as a separate process, executed one million summing iterations, and returned checksum 499999500000. Results include process startup, program loading, and execution.