ModelCap

Current AI model comparison

Exact identities · public evidence

Cohere

North Mini Code

Position

#129

Index

34.7

Estimated

3 reported rows against measured corpus ladders (61%); publisher-corpus-prior over 188 held-out anchors (39%); shrunk 12.4 toward the measured corpus

Comparison side 1

Mistral AI

Mistral Small 3.2 24B

Position

#127

Index

36.1

Measured

2 public benchmark observations across 2 boards

Comparison side 2

Output / 1M

Free

vs $0.20

Context

256K

vs 256K

Weights

Open weights

vs Open weights