Model comparison
v6-Finch-14B-HF vs gemma-4-E4B-it
At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 8,767,898,176. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 60.3× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| v6-Finch-14B-HF | gemma-4-E4B-it | |
|---|---|---|
| Parameters | 14.1B | 8.0B |
| Architecture | rwkv6 | gemma4 |
| Layers | 61 | 42 |
| Native context | — | 131,072 |
| Mixture of experts | no | no |
| Quantizations published | 20 | 36 |
| Smallest quantization | 5.10 GiB | 3.30 GiB |
| Q4_K_M | 8.17 GiB | 4.64 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | v6-Finch-14B-HF | gemma-4-E4B-it | Ratio |
|---|---|---|---|
| 4,096 | 3.81 GiB | 0.12 GiB | 30.98× |
| 8,192 | 7.63 GiB | 0.18 GiB | 42.90× |
| 16,384 | 15.25 GiB | 0.29 GiB | 53.12× |
| 32,768 | 30.50 GiB | 0.51 GiB | 60.29× |
| 65,536 | 61.00 GiB | 0.94 GiB | 64.66× |
| 131,072 | 122.00 GiB | 1.82 GiB | 67.09× |