How much VRAM does gemma-4-26B-A4B-it-Claude-Opus-Distill need?

Its published quantizations range from 11.38 GiB to 47.04 GiB in weights alone, before KV cache and runtime overhead.

gemma-4-26B-A4B-it-Claude-Opus-Distill on Radeon RX 6500 XT

Can I run gemma-4-26B-A4B-it-Claude-Opus-Distill on a Radeon RX 6500 XT?

Q: Can I run gemma-4-26B-A4B-it-Claude-Opus-Distill on a Radeon RX 6500 XT?

No. No published quantization of gemma-4-26B-A4B-it-Claude-Opus-Distill fits a Radeon RX 6500 XT at any context with q8_0 KV cache.

Not at these settings. No indexed quantization of gemma-4-26B-A4B-it-Claude-Opus-Distill fits Radeon RX 6500 XT at any context we compute, with q8_0 KV. The smallest shipped quantization is 11.38 GiB in weights alone, against 3.72 GiB usable. CPU offload can still run it, slowly.

From the file· weights summed from filesFrom the file· KV computed per layerPredicted· speed and compute buffer

KV cache dtypef16 q8_0 q4_0— 4 GB card, 3.72 GiB usable after overhead

Every quantization at every context

total memory required; green fits, red does not

Quant	Weights●	4K◐	8K◐	16K◐	32K◐	64K◐	128K◐
BF16	47.04 GiB	48.2	48.2	48.4	48.7	49.4	50.7
Q8_0	25.02 GiB	26.1	26.2	26.4	26.7	27.4	28.7
Q6_K	21.08 GiB	22.2	22.3	22.5	22.8	23.5	24.8
Q5_K_M	17.82 GiB	18.9	19.0	19.2	19.5	20.2	21.5
Q4_K_M	15.64 GiB	16.8	16.9	17.0	17.3	18.0	19.3
IQ4_NL	13.58 GiB	14.7	14.8	15.0	15.3	16.0	17.3
Q3_K_M	12.37 GiB	13.5	13.6	13.7	14.1	14.7	16.1
Q3_K_S	11.38 GiB	12.5	12.6	12.8	13.1	13.8	15.1

●From the file◐Predictedwhat these mean

Figures are GiB of total memory: weights plus KV cache plus compute buffer and backend overhead. Weights and KV are near-exact; the overhead term is modeled. Hover any cell for the breakdown.

Why other calculators disagree

A parameters × bits ÷ 8 estimate ignores two things that dominate at long context. First, the weights themselves are not the nominal rate — quantizations are mixtures, so the real file is consistently larger than the label implies. Second, most of this model's layers cache only a 1,024-token window rather than the full context.