Compare · 2026-07 cohort

Compare humanoids on Index scores.

Everything here uses the same numbers as the Index: one score across 8 axes and 40 sub-axes. For each difference we also show how much public evidence sits behind it. Where there is no evidence yet, we say so rather than scoring a zero.

Pick up to 4

Nothing selected yet

Pick two humanoids to compare.

You can also compare one humanoid against the cohort average or the cohort best. Both are worked out sub-axis by sub-axis from every published entry, not copied from any single robot.

All comparisons

Every pair in the July 2026 cohort

66 comparisons, each built from the published Index scores. They update on their own when the next month is signed off.

LevelWide apart

Each cell is the Capability gap between the two. “Thin” marks a pair where one side has fewer than 20 of the 40 sub-axes on the public record — the page still opens, it just says plainly how little there is to compare.