Model comparison
Qwen3.8-27B vs Qwen3-15B-A2B-Base
At Q4_K_M, Qwen3-15B-A2B-Base is the smaller download — 9,496,453,792 bytes against 16,810,714,336. At long context the gap widens: Qwen3-15B-A2B-Base's KV cache at 32K is 1.3× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| Qwen3.8-27B | Qwen3-15B-A2B-Base | |
|---|---|---|
| Parameters | 27.8B | 15.6B |
| Architecture | qwen35 | qwen3moe |
| Layers | 64 | 24 |
| Native context | 262,144 | 4,096 |
| Mixture of experts | no | yes, 128 experts |
| Quantizations published | 22 | 11 |
| Smallest quantization | 8.39 GiB | 5.41 GiB |
| Q4_K_M | 15.66 GiB | 8.84 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Qwen3.8-27B | Qwen3-15B-A2B-Base | Ratio |
|---|---|---|---|
| 4,096 | 0.25 GiB | 0.19 GiB | 1.33× |
| 8,192 | 0.50 GiB | 0.38 GiB | 1.33× |
| 16,384 | 1.00 GiB | 0.75 GiB | 1.33× |
| 32,768 | 2.00 GiB | 1.50 GiB | 1.33× |
| 65,536 | 4.00 GiB | 3.00 GiB | 1.33× |
| 131,072 | 8.00 GiB | 6.00 GiB | 1.33× |