Can I run this LLM? Real file sizes, live from Hugging Face.
All models › gemma-4-26B-A4B-it

gemma-4-26B-A4B-it requirements — can you run it?

~25.3B parameters · 20 quantizations measured · updated 2026-07-17 · source: unsloth/gemma-4-26B-A4B-it-GGUF

Run gemma-4-26B-A4B-it (UD-IQ2_XXS) — copy-paste:
# llama.cpp (auto-downloads the GGUF)
llama-cli -hf unsloth/gemma-4-26B-A4B-it-GGUF:UD-IQ2_XXS

# Ollama (pulls straight from Hugging Face)
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:UD-IQ2_XXS

# LM Studio: search "unsloth/gemma-4-26B-A4B-it-GGUF" in the model browser

Commands load the exact quant measured on this page.

Every quantization: real file size vs what you actually need

QuantFile (measured)VRAM/RAM needed (8K ctx)Runs on (first fit)
UD-IQ2_XXS9.9 GB12.9 GBRTX 4060 Ti 16GB
UD-IQ2_M10.0 GB13.0 GBRTX 4060 Ti 16GB
UD-Q2_K_XL10.5 GB13.6 GBRTX 4060 Ti 16GB
UD-IQ3_S11.3 GB14.3 GBRTX 4060 Ti 16GB
UD-IQ3_XXS11.4 GB14.4 GBRTX 4060 Ti 16GB
UD-Q3_K_M12.7 GB15.8 GBRTX 3090 24GB
UD-Q3_K_XL12.9 GB16.0 GBRTX 3090 24GB
UD-IQ4_XS13.6 GB16.7 GBRTX 3090 24GB
UD-IQ4_NL13.6 GB16.7 GBRTX 3090 24GB
UD-Q4_K_S16.5 GB19.8 GBRTX 3090 24GB
UD-Q4_K_M16.9 GB20.3 GBRTX 3090 24GB
UD-Q4_K_XL17.0 GB20.4 GBRTX 3090 24GB
UD-Q5_K_S18.9 GB22.4 GBRTX 5090 32GB
UD-Q5_K_M21.2 GB24.9 GBRTX 5090 32GB
UD-Q5_K_XL21.2 GB24.9 GBRTX 5090 32GB
UD-Q6_K23.2 GB27.0 GBRTX 5090 32GB
UD-Q6_K_XL23.3 GB27.2 GBRTX 5090 32GB
Q8_026.9 GB31.0 GBRTX PRO 6000 96GB
UD-Q8_K_XL27.6 GB31.9 GBRTX PRO 6000 96GB
BF1650.5 GB56.6 GBRTX PRO 6000 96GB

KV cache computed exactly from the model config (GQA formula, 8K context). MoE models keep all experts in memory — total size counts, not just active params. Longer context grows the KV cache — use the selector above.

Can I run gemma-4-26B-A4B-it on my GPU? (UD-IQ2_XXS recommended quant)

HardwareMemoryVerdictEst. speed
RTX 3060 12GB12 GBno
RTX 4060 Ti 16GB16 GBRUNS~22 tok/s
RTX 3090 24GB24 GBRUNS~71 tok/s
RTX 4090 24GB24 GBRUNS~76 tok/s
RTX 5090 32GB32 GBRUNS~135 tok/s
RTX PRO 6000 96GB96 GBRUNS~135 tok/s
Mac 16GB unified16 GBRUNS~8 tok/s
Mac 32GB unified32 GBRUNS~11 tok/s
Mac 64GB unified64 GBRUNS~21 tok/s
Mac 128GB unified128 GBRUNS~30 tok/s
Mac 256GB unified256 GBRUNS~41 tok/s
Mac 512GB unified512 GBRUNS~62 tok/s
32GB system RAM (CPU)32 GBRUNS~4 tok/s
64GB system RAM (CPU)64 GBRUNS~4 tok/s
128GB system RAM (CPU)128 GBRUNS~6 tok/s
256GB system RAM (CPU)256 GBRUNS~6 tok/s
1TB server RAM (CPU)1024 GBRUNS~15 tok/s
2TB server RAM (CPU)2048 GBRUNS~15 tok/s

Speed = decode tok/s estimated from memory bandwidth ÷ measured file size (× MoE active share). Real numbers vary ±30% by runtime and settings.

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

The honest answer: 12.9 GB at the UD-IQ2_XXS quant with 8K context — that is the measured file size (9.9 GB) plus KV cache and runtime overhead. The smallest published build needs 12.9 GB; the lossless one needs 56.6 GB.
Cheapest hardware that runs it: New RTX 4060 Ti 16GB (~$420, used market) — see which GPU to buy.

Can I run it with CPU offload?

Yes, if combined RAM+VRAM ≥ file size — but generation speed is limited by memory bandwidth. Expect roughly: DDR5 dual channel ~10-30 tok/s for small MoE actives, single digits for big ones, 0.1-1 tok/s when experts page from disk. CPU-only is fine for batch jobs, painful for chat.

Why do other sites show different numbers?

Most pages compute weights from a formula (params × bits/8). We use the actual file sizes published in GGUF repos, which include the embedding table, unquantized tensors, and container overhead — that is why our numbers can differ from a naive calculation.

FAQ

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

The measured answer: 12.9 GB at the UD-IQ2_XXS quant with 8K context. The smallest published build needs 12.9 GB; the lossless (F16/BF16) build needs 56.6 GB. These are real GGUF file sizes from Hugging Face, not formula estimates.

Can I run gemma-4-26B-A4B-it on a 24GB GPU (RTX 3090/4090)?

Yes — at UD-IQ2_XXS (12.9 GB). Quants that fit 24GB: UD-IQ2_XXS, UD-IQ2_M, UD-Q2_K_XL, UD-IQ3_S, UD-IQ3_XXS, UD-Q3_K_M, UD-Q3_K_XL, UD-IQ4_XS, UD-IQ4_NL, UD-Q4_K_S, UD-Q4_K_M, UD-Q4_K_XL, UD-Q5_K_S.

How fast will gemma-4-26B-A4B-it run?

Decode speed is memory-bandwidth-bound. At UD-IQ2_XXS: roughly 76 tok/s on an RTX 4090, 71 tok/s on an RTX 3090, 30 tok/s on an M-series Mac with 128GB, and 3.8 tok/s on CPU with dual-channel DDR4 — estimates from measured file size, MoE active-parameter share and memory bandwidth.

Can I run gemma-4-26B-A4B-it on CPU with system RAM?

Yes, if your usable RAM is at least the file size. At UD-IQ2_XXS you need ~12.9 GB of RAM. CPU generation is memory-bandwidth-bound: expect single-digit tok/s for large MoE models, 10-30 tok/s for small active params, and below 1 tok/s when experts page from disk.

Why do VRAM numbers for gemma-4-26B-A4B-it differ between sites?

Most sites compute weights from a formula (params x bits/8). ModelFit uses the actual GGUF file sizes published on Hugging Face, which include the embedding table, unquantized tensors, and container overhead — so our numbers reflect what you really download and load.

More models

gemma-4-E4B-itmeasured requirementsGemma-4-E4B-Uncensored-HauhauCS-Aggressivemeasured requirementsgemma-4-12b-itmeasured requirementsgemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2measured requirementsgemma-4-E4B-itmeasured requirementsgemma-3-4b-itmeasured requirementsgemma-3-4b-itmeasured requirementsgemma-3-12b-itmeasured requirements
Share this page: 𝕏 Post Reddit Hacker News Telegram WhatsApp More…