batiai/DeepSeek-V4-Pro-GGUF
Text Generation β’ 1.6T β’ Updated β’ 259 β’ 2
π Frontier MoE β 128Bβ1T
Largest open-weight LLMs, BatiAI-quantized. Mac-runnable from M4 Max 128GB to Mac Studio M3 Ultra 512GB.
-
-
batiai/Kimi-K2.6-GGUF
Text Generation β’ 1T β’ Updated β’ 120 -
batiai/GLM-5.1-GGUF
Text Generation β’ 754B β’ Updated β’ 14
batiai/DeepSeek-V4-Flash-GGUF
Text Generation β’ 284B β’ Updated β’ 5.95k β’ 12Note 284B-A13B MoE β’ CSA+HCA hybrid attention β’ SWE-Bench Pro top tier β’ via batiai/bati.cpp
-
batiai/MiniMax-M2.7-GGUF
Text Generation β’ 229B β’ Updated β’ 26
batiai/Mistral-Medium-3.5-128B-GGUF
Text Generation β’ 125B β’ Updated β’ 172Note 128B Dense β’ SWE-Bench Verified 77.6% β’ Modified MIT β’ measured 6.8 t/s on M4 Max IQ3
-
batiai/Llama-4-Scout-17B-16E-Instruct-GGUF
Image-Text-to-Text β’ 108B β’ Updated β’ 638