all the models — AI benchmark observatory
← Benchmarks

COLLIE

COLLIE is a grammar-based framework for systematic construction of constrained text generation tasks. It allows specification of rich, compositional constraints across diverse generation levels and modeling challenges including language understanding, logical reasoning, and semantic planning. The COLLIE-v1 dataset contains 2,080 instances across 13 constraint structures.

id collie · max 1 · 11 models reported

#ModelScore
1GPT-5
OpenAI
0.99
2o3-mini
OpenAI
0.99
3o3
OpenAI
0.98
4Mistral Medium 3.5
Mistral AI · open
0.96
COLLIE Leaderboard · all the models