Chart · milliseconds
filter amount > 500, count.
neo 0.018 ms in the language study, 0.018 ms in the library run. Closest: Zig at 1.67×. Farthest: Spark at 2,049×.
Embed anywhere.
<iframe src="https://n0.is/embed/chart/filter" title="filter amount > 500, count" width="100%" height="520" loading="lazy" sandbox="allow-scripts" referrerpolicy="no-referrer" style="border:0;border-radius:12px"></iframe>Every engine's multiple of neo
- neo1×
- Zig1.67× · 0.030 vs 0.018
- Rust2.06× · 0.037 vs 0.018
- Swift2.33× · 0.042 vs 0.018
- Go3.56× · 0.064 vs 0.018
- Polars9.94× · 0.179 vs 0.018
- pandas18.2× · 0.328 vs 0.018
- DuckDB33.5× · 0.602 vs 0.018
- Spark2,049× · 36.9 vs 0.018
Bars to scale, from zero to the slowest engine. The gray pair is the engine's measurement and neo's in the same run, in ms. The languages and the libraries were measured in separate runs, so neo has a number in each; every multiple is against neo in its own run.
What this measures
Count the rows whose amount is over 500: a million comparisons on a typed column, the answer one integer. Each library's count is written the way its users write it.
What neo is held to
The count every other engine prints, under the JavaScript node's rule that NaN never matches. What comes out is a selection the next kernel reads from, not a copy of the rows.
Each side, this kernel only
Zig1.67× neo · neo vs Zig
Written in-house; it stays with the engine. How it was run and checked is on its comparison page.
Rust2.06× neo · neo vs Rust
Written in-house; it stays with the engine. How it was run and checked is on its comparison page.
Swift2.33× neo · neo vs Swift
Written in-house; it stays with the engine. How it was run and checked is on its comparison page.
Go3.56× neo · neo vs Go
Written in-house; it stays with the engine. How it was run and checked is on its comparison page.
Polars9.94× neo · neo vs Polars
timed("filter", lambda: df.lazy().filter(pred).select(pl.len()).collect().item())pandas18.2× neo · neo vs pandas
timed("filter", lambda: int((amount > 500).sum()))DuckDB33.5× neo · neo vs DuckDB
timed("filter", lambda: con.execute("SELECT count(*) FROM t WHERE amount > 500").fetchone())Spark2,049× neo · neo vs Spark
timed("filter", lambda: df.filter(pred).count())neo · confidential
neo's kernels, the ports written in-house to beat them, and the techniques behind both stay with the engine. Nothing about how the numbers were made is withheld: the method, the checks every engine had to pass, every number and every place neo is behind are on this page. The engine itself we keep confidential, because we don't want it getting into the wrong hands. The binaries run on request and reproduce every number here.
Measured on Apple M5 Max, 18 cores, 26.4.1, 2026-10-11 and 2026-10-15. Every number here is also on the comparison page of its engine, with how it was taken and the caveats.