Model comparison
gemma-4-E4B-it-assistant vs SmolLM2-135M-Instruct
At Q4_K_M, gemma-4-E4B-it-assistant is the smaller download — 77,231,200 bytes against 105,454,144. At long context the gap widens: gemma-4-E4B-it-assistant's KV cache at 32K is 10.3× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| gemma-4-E4B-it-assistant | SmolLM2-135M-Instruct | |
|---|---|---|
| Parameters | 79M | 135M |
| Architecture | gemma4-assistant | llama |
| Layers | 4 | 30 |
| Native context | 131,072 | 8,192 |
| Mixture of experts | no | no |
| Quantizations published | 9 | 41 |
| Smallest quantization | 0.07 GiB | 0.08 GiB |
| Q4_K_M | 0.07 GiB | 0.10 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | gemma-4-E4B-it-assistant | SmolLM2-135M-Instruct | Ratio |
|---|---|---|---|
| 4,096 | 0.01 GiB | 0.09 GiB | 6.43× |
| 8,192 | 0.02 GiB | 0.18 GiB | 8.18× |
| 16,384 | 0.04 GiB | 0.35 GiB | 9.47× |
| 32,768 | 0.07 GiB | 0.70 GiB | 10.29× |
| 65,536 | 0.13 GiB | 1.41 GiB | 10.75× |
| 131,072 | 0.26 GiB | 2.81 GiB | 10.99× |