Model comparison
Ling-mini-2.0 vs gemma-4-12B-it-qat-q4_0-unquantized
These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: Ling-mini-2.0's KV cache at 32K is 2.0× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| Ling-mini-2.0 | gemma-4-12B-it-qat-q4_0-unquantized | |
|---|---|---|
| Parameters | 16.3B | 12.0B |
| Architecture | bailingmoe2 | gemma4 |
| Layers | 20 | 48 |
| Native context | 32,768 | 262,144 |
| Mixture of experts | yes, 256 experts | no |
| Quantizations published | 30 | 2 |
| Smallest quantization | 4.34 GiB | 6.50 GiB |
| Q4_K_M | 9.23 GiB | — |
| Licence | — | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Ling-mini-2.0 | gemma-4-12B-it-qat-q4_0-unquantized | Ratio |
|---|---|---|---|
| 4,096 | 0.16 GiB | 0.72 GiB | 4.60× |
| 8,192 | 0.31 GiB | 0.97 GiB | 3.10× |
| 16,384 | 0.63 GiB | 1.47 GiB | 2.35× |
| 32,768 | 1.25 GiB | 2.47 GiB | 1.98× |
| 65,536 | 2.50 GiB | 4.47 GiB | 1.79× |
| 131,072 | 5.00 GiB | 8.47 GiB | 1.69× |