Model comparison

Dragon-1-0.5B vs Qwen3-1.7B

At Q4_K_M, Dragon-1-0.5B is the smaller download — 397,523,744 bytes against 1,107,409,472. At long context the gap widens: Dragon-1-0.5B's KV cache at 32K is 9.3× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Dragon-1-0.5BQwen3-1.7B
Parameters494M2.0B
Architectureqwen2qwen3
Layers2428
Native context32,76840,960
Mixture of expertsnono
Quantizations published3648
Smallest quantization0.29 GiB0.50 GiB
Q4_K_M0.37 GiB1.03 GiB
Licenceapache-2.0—

KV cache by context

the term that decides long-context viability
ContextDragon-1-0.5BQwen3-1.7BRatio
4,0960.05 GiB0.44 GiB9.33×
8,1920.09 GiB0.88 GiB9.33×
16,3840.19 GiB1.75 GiB9.33×
32,7680.38 GiB3.50 GiB9.33×
65,5360.75 GiB7.00 GiB9.33×
131,0721.50 GiB14.00 GiB9.33×