A list of (newer) LLMs that adhere to full transparancy and reproducibility by openly sharing their full training data and training recipes.
-
IFM/K2-Horizon-MoVA-36B-A4B-GGUF
Text Generation • 37B • Updated • 26.7k • 131 -
swiss-ai/Apertus-v1.5-70B
Image-Text-to-Text • 72B • Updated • 17.7k • • 105 -
swiss-ai/Apertus-v1.5-8B
Image-Text-to-Text • 9B • Updated • 394k • 111 -
allenai/Olmo-3.1-32B-Think
Text Generation • 32B • Updated • 5.25k • 115