Mixed-precision GGUF quantizations of moonshotai/Kimi-K2.6 from the RAM pipeline (per-tensor bit allocation via sensitivity probing).
AI & ML interests
Model Quantization
Recent Activity
Organization Card
Smaller. Smarter. Sovereign.
Making frontier models run anywhere
We publish high-quality quantized models for Apple Silicon and GGUF. Our models use a proprietary optimisation method that delivers superior quality at your target memory budget.
Browse our models, or connect with us below.
models 58
baa-ai/Kimi-K2.6-RAM-GGUF
Text Generation • 1T • Updated • 305
baa-ai/Kimi-K2.6-RAM-447GB-GGUF
Text Generation • 1T • Updated • 602
baa-ai/Kimi-K2.6-RAM-344GB-GGUF
Text Generation • 1T • Updated • 200
baa-ai/MiniMax-M2.7-RAM-117GB-MLX
229B • Updated • 163
baa-ai/MiniMax-M2.7-RAM-130GB-MLX
229B • Updated • 102
baa-ai/Llama-4-Scout-17B-16E-Instruct-RAM-137GB-GGUF
Text Generation • 108B • Updated • 179
baa-ai/Llama-4-Scout-17B-16E-Instruct-RAM-105GB-GGUF
Text Generation • 108B • Updated • 171
baa-ai/Qwen3.5-122B-A10B-RAM-164GB-GGUF
Text Generation • 122B • Updated • 339
baa-ai/Qwen3.5-122B-A10B-RAM-94GB-GGUF
Text Generation • 122B • Updated • 161
baa-ai/Gemma-4-31B-it-RAM-26GB-GGUF
Text Generation • 31B • Updated • 686
datasets 0
None public yet