Is Qwen3-Coder-Next-DFlash or embeddinggemma-300m smaller?

Both publish quantizations; compare the table for the exact sizes.

Which uses less memory at long context?

embeddinggemma-300m — its KV cache at 32K context is 3.5x smaller, because of how its attention layers cache.

Model comparison

Qwen3-Coder-Next-DFlash vs embeddinggemma-300m

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: embeddinggemma-300m's KV cache at 32K is 3.5× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3-Coder-Next-DFlash	embeddinggemma-300m
Parameters	474M	303M
Architecture	dflash	gemma-embedding
Layers	8	24
Native context	262,144	2,048
Mixture of experts	no	no
Quantizations published	7	10
Smallest quantization	0.25 GiB	0.26 GiB
Q4_K_M	0.27 GiB	—
Licence	—	—

KV cache by context

the term that decides long-context viability

Context	Qwen3-Coder-Next-DFlash	embeddinggemma-300m	Ratio
4,096	0.06 GiB	0.04 GiB	1.78×
8,192	0.13 GiB	0.05 GiB	2.46×
16,384	0.25 GiB	0.08 GiB	3.05×
32,768	0.50 GiB	0.14 GiB	3.46×
65,536	1.00 GiB	0.27 GiB	3.71×
131,072	2.00 GiB	0.52 GiB	3.85×