Model comparison

Trinity-Nano-Preview vs Qwen3-8B

At Q4_K_M, Trinity-Nano-Preview is the smaller download — 3,786,957,088 bytes against 5,027,783,488. At long context the gap widens: Trinity-Nano-Preview's KV cache at 32K is 8.3× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Trinity-Nano-PreviewQwen3-8B
Parameters6.1B8.2B
Architectureafmoeqwen3
Layers5636
Native context131,07240,960
Mixture of expertsyes, 128 expertsno
Quantizations published5051
Smallest quantization1.94 GiB2.12 GiB
Q4_K_M3.53 GiB4.68 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextTrinity-Nano-PreviewQwen3-8BRatio
4,0960.16 GiB0.56 GiB3.58×
8,1920.21 GiB1.13 GiB5.31×
16,3840.32 GiB2.25 GiB7.00×
32,7680.54 GiB4.50 GiB8.33×
65,5360.98 GiB9.00 GiB9.21×
131,0721.85 GiB18.00 GiB9.72×