ModelCap

Current AI model comparison

Exact identities · public evidence

Kwaipilot

KAT Coder V2.5 Dev

Position

#37

Index

65.5

Estimated

global-corpus-prior over 194 held-out anchors (23%); launch card against 2 resolved peers on 4 rows (77%); exceeds every named peer on 4 of 4 rows; optimism haircut 0 from cross-lab-probe; shrunk 3.9 toward the measured corpus

Comparison side 1

Qwen

Qwen3.5 397B A17B

Position

#42

Index

63.4

Measured

3 public benchmark observations across 3 boards

Comparison side 2

Output / 1M

Unavailable

vs $3.50

Context

vs 262K

Weights

Open weights

vs Open weights