Model comparison
gemma-4-26B-A4B-it vs gemma-4-E4B-it-qat-q4_0-unquantized
These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: gemma-4-E4B-it-qat-q4_0-unquantized's KV cache at 32K is 3.1× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| gemma-4-26B-A4B-it | gemma-4-E4B-it-qat-q4_0-unquantized | |
|---|---|---|
| Parameters | 26.5B | 7.9B |
| Architecture | gemma4 | gemma4 |
| Layers | 30 | 42 |
| Native context | 262,144 | 131,072 |
| Mixture of experts | yes, 128 experts | no |
| Quantizations published | 46 | 1 |
| Smallest quantization | 8.99 GiB | 4.80 GiB |
| Q4_K_M | 15.64 GiB | — |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | gemma-4-26B-A4B-it | gemma-4-E4B-it-qat-q4_0-unquantized | Ratio |
|---|---|---|---|
| 4,096 | 0.45 GiB | 0.12 GiB | 3.65× |
| 8,192 | 0.61 GiB | 0.18 GiB | 3.41× |
| 16,384 | 0.92 GiB | 0.29 GiB | 3.20× |
| 32,768 | 1.54 GiB | 0.51 GiB | 3.05× |
| 65,536 | 2.79 GiB | 0.94 GiB | 2.96× |
| 131,072 | 5.29 GiB | 1.82 GiB | 2.91× |