Hybrid GGUF quants created via an evolutionary quant algorithm. Want the best TPS? Lowest precision loss? Smallest file size? Welcome to MagicQuant!
-
magiccodingman/Granite-4.0-H-350M-Unsloth-MagicQuant-Hybrid-GGUF
Text Generation • 0.3B • Updated • 88 • 1 -
magiccodingman/Qwen3-4B-Thinking-2507-Unsloth-MagicQuant-Hybrid-GGUF
Text Generation • 4B • Updated • 334 • 1 -
magiccodingman/Qwen3-4B-Instruct-2507-Unsloth-MagicQuant-Hybrid-GGUF
Text Generation • 4B • Updated • 580 • 6 -
magiccodingman/Qwen3-30B-A3B-Thinking-2507-unsloth-MagicQuant-Hybrid-GGUF
Text Generation • 31B • Updated • 439 • 5