Synced from Hugging Face — 2026-09-14
How much VRAM does gemma-2-27b-it-GGUF need?
Exact on-disk weight sizes below are pulled directly from Hugging Face. Total VRAM figures also include an estimated KV cache and runtime overhead — see the breakdown per quantization, and the calculator below to adjust context length.
By quantization
gemma-2-27b-it-IQ2_M
8.75 GB exact, on disk
~13.06 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ2_S
8.06 GB exact, on disk
~12.30 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ2_XS
7.82 GB exact, on disk
~12.04 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ3_M
11.60 GB exact, on disk
~16.20 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ3_XS
10.76 GB exact, on disk
~15.27 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ3_XXS
10.01 GB exact, on disk
~14.45 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-IQ4_XS
13.80 GB exact, on disk
~18.62 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q2_K
9.73 GB exact, on disk
~14.14 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q2_K_L
10.00 GB exact, on disk
~14.44 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q3_K_L
13.52 GB exact, on disk
~18.31 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q3_K_M
12.50 GB exact, on disk
~17.19 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q3_K_S
11.33 GB exact, on disk
~15.90 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q3_K_XL
13.79 GB exact, on disk
~18.61 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q4_K_L
15.77 GB exact, on disk
~20.78 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q4_K_M
15.50 GB exact, on disk
~20.49 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q4_K_S
14.66 GB exact, on disk
~19.56 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q5_K_L
18.34 GB exact, on disk
~23.61 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q5_K_M
18.08 GB exact, on disk
~23.33 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q5_K_S
17.59 GB exact, on disk
~22.79 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q6_K
20.81 GB exact, on disk
~26.33 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q6_K_L
21.08 GB exact, on disk
~26.63 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q8_0
26.95 GB exact, on disk
~33.08 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-Q8_0_L
27.98 GB exact, on disk
~34.22 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00001-of-00003
36.89 GB exact, on disk
~44.02 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00002-of-00003
37.13 GB exact, on disk
~44.28 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00003-of-00003
27.42 GB exact, on disk
~33.60 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32/gemma-2-27b-it-f32-00001-of-00003
36.89 GB exact, on disk
~44.02 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32/gemma-2-27b-it-f32-00002-of-00003
37.13 GB exact, on disk
~44.28 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
gemma-2-27b-it-f32/gemma-2-27b-it-f32-00003-of-00003
27.42 GB exact, on disk
~33.60 GB total estimated VRAM at 4,096 tokens context (KV cache + overhead estimated, not exact)
VRAM calculator
Estimate total VRAM for a given model size, quantization, and context length. Estimates only — see the note at the bottom.
- Model weights
- —
- KV cache
- —
- Runtime overhead (10%)
- —
- Total VRAM
- —
Fits on:
This won't fit on local hardware
Based on the estimate above, this configuration exceeds every local GPU profile listed. Renting cloud GPU capacity is the practical option for a workload this size.
RunPod
On-demand GPU pods, billed by the minute.
Lambda Labs
Reserved and on-demand cloud GPU instances.
Disclosure: We may earn a commission from cloud providers if you spin up an instance through these links, at no extra cost to you. This never affects the VRAM math above — see /methodology.
Live changes
- No changes yet — adjust a control above.
Generic estimate: weight size
is bits-per-weight × parameter count; KV cache uses a reference
architecture for the selected size bucket (7B/13B/70B: real published
models; 120B: extrapolated, no real model exists at that exact
size). Model-specific (HF-synced):
weight size is the exact .gguf file size fetched from
Hugging Face — no estimation. KV cache is still estimated:
this schema doesn't carry layer count, head count, or head dimension,
so the calculator parses an approximate parameter count from the
model's title (e.g. "8B") and reuses the nearest size bucket's
reference architecture for the KV math, same caveats as the generic
mode. If no parameter count can be parsed from the title, it falls
back to the 7B architecture and says so next to the quantization
dropdown. Batch size is fixed at 1 in both modes. Treat all figures
here as a starting estimate, not a guarantee. Full formula and known
limitations: /methodology.
FAQ
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ2_M?
The gemma-2-27b-it-IQ2_M quantization is 8.75 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 13.06 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ2_S?
The gemma-2-27b-it-IQ2_S quantization is 8.06 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 12.30 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ2_XS?
The gemma-2-27b-it-IQ2_XS quantization is 7.82 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 12.04 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ3_M?
The gemma-2-27b-it-IQ3_M quantization is 11.60 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 16.20 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ3_XS?
The gemma-2-27b-it-IQ3_XS quantization is 10.76 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 15.27 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ3_XXS?
The gemma-2-27b-it-IQ3_XXS quantization is 10.01 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 14.45 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-IQ4_XS?
The gemma-2-27b-it-IQ4_XS quantization is 13.80 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 18.62 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q2_K?
The gemma-2-27b-it-Q2_K quantization is 9.73 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 14.14 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q2_K_L?
The gemma-2-27b-it-Q2_K_L quantization is 10.00 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 14.44 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q3_K_L?
The gemma-2-27b-it-Q3_K_L quantization is 13.52 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 18.31 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q3_K_M?
The gemma-2-27b-it-Q3_K_M quantization is 12.50 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 17.19 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q3_K_S?
The gemma-2-27b-it-Q3_K_S quantization is 11.33 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 15.90 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q3_K_XL?
The gemma-2-27b-it-Q3_K_XL quantization is 13.79 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 18.61 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q4_K_L?
The gemma-2-27b-it-Q4_K_L quantization is 15.77 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 20.78 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q4_K_M?
The gemma-2-27b-it-Q4_K_M quantization is 15.50 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 20.49 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q4_K_S?
The gemma-2-27b-it-Q4_K_S quantization is 14.66 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 19.56 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q5_K_L?
The gemma-2-27b-it-Q5_K_L quantization is 18.34 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 23.61 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q5_K_M?
The gemma-2-27b-it-Q5_K_M quantization is 18.08 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 23.33 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q5_K_S?
The gemma-2-27b-it-Q5_K_S quantization is 17.59 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 22.79 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q6_K?
The gemma-2-27b-it-Q6_K quantization is 20.81 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 26.33 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q6_K_L?
The gemma-2-27b-it-Q6_K_L quantization is 21.08 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 26.63 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q8_0?
The gemma-2-27b-it-Q8_0 quantization is 26.95 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 33.08 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-Q8_0_L?
The gemma-2-27b-it-Q8_0_L quantization is 27.98 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 34.22 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00001-of-00003?
The gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00001-of-00003 quantization is 36.89 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 44.02 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00002-of-00003?
The gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00002-of-00003 quantization is 37.13 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 44.28 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00003-of-00003?
The gemma-2-27b-it-f32.gguf/gemma-2-27b-it-f32-00003-of-00003 quantization is 27.42 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 33.60 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32/gemma-2-27b-it-f32-00001-of-00003?
The gemma-2-27b-it-f32/gemma-2-27b-it-f32-00001-of-00003 quantization is 36.89 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 44.02 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32/gemma-2-27b-it-f32-00002-of-00003?
The gemma-2-27b-it-f32/gemma-2-27b-it-f32-00002-of-00003 quantization is 37.13 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 44.28 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.
How much VRAM does gemma-2-27b-it-GGUF need at gemma-2-27b-it-f32/gemma-2-27b-it-f32-00003-of-00003?
The gemma-2-27b-it-f32/gemma-2-27b-it-f32-00003-of-00003 quantization is 27.42 GB on disk (exact, synced from Hugging Face). At a 4,096-token context, using the 13B reference architecture for the KV cache estimate, total estimated VRAM is approximately 33.60 GB. This total includes a KV cache estimate and a 10% runtime overhead margin — only the weight size is exact.