⚡ DeepSeek Coder 1.3B — CPU Free Tier

llama.cpp + GGUF Q4_K_M · 2 vCPU / 16 GB · ~15-25 t/s

64 1024
0 1
0.5 1
1 80
1 1.5