Model comparison

LFM2-24B-A2B vs Qwen3.6-27B

At Q4_K_M, LFM2-24B-A2B is the smaller download — 14,415,473,952 bytes against 16,547,397,888. At long context the gap widens: LFM2-24B-A2B's KV cache at 32K is 3.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

LFM2-24B-A2BQwen3.6-27B
Parameters23.8B27.8B
Architecturelfm2moeqwen35
Layers4064
Native context128,000262,144
Mixture of expertsyes, 64 expertsno
Quantizations published3340
Smallest quantization5.35 GiB8.74 GiB
Q4_K_M13.43 GiB15.41 GiB
Licenceotherapache-2.0

KV cache by context

the term that decides long-context viability
ContextLFM2-24B-A2BQwen3.6-27BRatio
4,0960.08 GiB0.25 GiB3.20×
8,1920.16 GiB0.50 GiB3.20×
16,3840.31 GiB1.00 GiB3.20×
32,7680.63 GiB2.00 GiB3.20×
65,5361.25 GiB4.00 GiB3.20×
131,0722.50 GiB8.00 GiB3.20×