Model comparison
embeddinggemma-300m vs gemma-3-1b-it
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| embeddinggemma-300m | gemma-3-1b-it | |
|---|---|---|
| Parameters | 303M | 1000M |
| Architecture | gemma-embedding | gemma3 |
| Layers | 24 | 26 |
| Native context | 2,048 | 32,768 |
| Mixture of experts | no | no |
| Quantizations published | 10 | 28 |
| Smallest quantization | 0.26 GiB | 0.52 GiB |
| Q4_K_M | — | 0.75 GiB |
| Licence | — | gemma |
KV cache by context
the term that decides long-context viability
| Context | embeddinggemma-300m | gemma-3-1b-it | Ratio |
|---|---|---|---|
| 4,096 | 0.04 GiB | 0.04 GiB | 1.06× |
| 8,192 | 0.05 GiB | 0.05 GiB | 1.04× |
| 16,384 | 0.08 GiB | 0.08 GiB | 1.02× |
| 32,768 | 0.14 GiB | 0.15 GiB | 1.01× |
| 65,536 | 0.27 GiB | 0.27 GiB | 1.01× |
| 131,072 | 0.52 GiB | 0.52 GiB | 1.00× |