aquif-ai/aquif-3.5-Max-1205
Text Generation
•
42B
•
Updated
•
127
•
8
Flagship models
Note new, thinking-only upgrade to the aquif-3.5-Max focused on coding and math performance.
Note our biggest model with 42.4B total, 3.3B active parameters and 1M context, made for hard coding and math tasks.
Note medium-sized 30.5B total, 3.3B active model with 1M context for general tasks and coding on 16GB+ VRAM cards.
Note small 1.7B model with 40K context for lightweight tasks and edge devices; outperforms aquif-3.5-3B and Ministral.