Model comparison
L3-DARKEST-PLANET-16.5B vs Qwen3.6-27B
At Q4_K_M, L3-DARKEST-PLANET-16.5B is the smaller download — 10,207,765,536 bytes against 16,547,397,888. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 4.4× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| L3-DARKEST-PLANET-16.5B | Qwen3.6-27B | |
|---|---|---|
| Parameters | 16.5B | 27.8B |
| Architecture | llama | qwen35 |
| Layers | 71 | 64 |
| Native context | 8,192 | 262,144 |
| Mixture of experts | no | no |
| Quantizations published | 10 | 40 |
| Smallest quantization | 7.00 GiB | 8.74 GiB |
| Q4_K_M | 9.51 GiB | 15.41 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | L3-DARKEST-PLANET-16.5B | Qwen3.6-27B | Ratio |
|---|---|---|---|
| 4,096 | 1.11 GiB | 0.25 GiB | 4.43× |
| 8,192 | 2.22 GiB | 0.50 GiB | 4.44× |
| 16,384 | 4.44 GiB | 1.00 GiB | 4.44× |
| 32,768 | 8.88 GiB | 2.00 GiB | 4.44× |
| 65,536 | 17.75 GiB | 4.00 GiB | 4.44× |
| 131,072 | 35.50 GiB | 8.00 GiB | 4.44× |