ModelCap

Current AI model comparison

Exact identities · public evidence

Cohere

North Mini Code

Position

#123

Index

33.9

Estimated

3 reported rows against measured corpus ladders (59%); publisher-corpus-prior over 183 held-out anchors (41%); shrunk 13.5 toward the measured corpus

Comparison side 1

IBM Granite

granite 4.2 30b

Position

#131

Index

32.5

Measured

2 public benchmark observations across 2 boards

Comparison side 2

Output / 1M

Free

vs Unavailable

Context

256K

vs 131K

Weights

Open weights

vs Open weights