ModelCap

Current AI model comparison

Exact identities · public evidence

Cohere

North Mini Code

Position

#135

Index

33.1

Estimated

3 reported rows against measured corpus ladders (56%); publisher-corpus-prior over 185 held-out anchors (44%); shrunk 14.3 toward the measured corpus

Comparison side 1

Microsoft

FrogBoss 32B 2510

Position

#134

Index

33.3

Inherited

finetune of qwen/qwen3-32b · finetune (100%)

Comparison side 2

Output / 1M

Free

vs Unavailable

Context

256K

vs —

Weights

Open weights

vs Open weights