How much VRAM to run Gemma 4 26B-A4B?

About 19 GB atQ4_K_M with an 8K context — fits a RX 7900 XTX. Full breakdown below, or check your exact hardware.

Gemma 4 26B-A4B VRAM by quantisation

QuantisationWeightsTotal (8K ctx)Fits on
Q4_K_M15.9 GB19.0 GBRX 7900 XTX, RTX 4090
Q5_K_M18.5 GB22.0 GBRX 7900 XTX, RTX 4090
Q6_K21.3 GB25.1 GBRTX 5090, Radeon AI PRO R9700
Q8_027.7 GB32.1 GBApple Silicon 48GB unified, RTX 6000 Ada
FP16 / BF1652.0 GB58.8 GBA100 80GB, RTX PRO 6000 Blackwell

Check your hardware

About Gemma 4 26B-A4B

Gemma 4 26B-A4B is Google's 26B-parameter model released in April 2026, with a 256K-token context window. It is a mixture-of-experts model: all 26B parameters must sit in memory, but only ~4B are active per token, which is what makes it fast for its size. It uses a hybrid-attention design — only a fraction of its layers cache the full context, so long conversations cost far less VRAM than a classic dense model: its KV cache is about 0.5 GB at an 8K context, 5.6 GB at 128K, and 10.9 GB at the full 256K window (FP16 cache).

For most people Q4_K_M is the sweet spot — the most popular quality/size trade-off — while Q8 is near-lossless if you have the memory. Totals above include the KV cache and a realistic framework overhead, so they are what you should expect to see in practice rather than just the download size. Weight sizes are calibrated against real GGUF files — see themethodology.

Frequently asked questions

How much VRAM does Gemma 4 26B-A4B need?

At Q4_K_M with an 8K context, Gemma 4 26B-A4B needs about 19 GB (weights 16 GB + KV cache + overhead). The smallest common hardware that fits is a RX 7900 XTX.

Can an RTX 4090 (24GB) run Gemma 4 26B-A4B?

Yes. An RTX 4090's 24 GB runs Gemma 4 26B-A4B at Q5_K_M (about 22 GB at 8K context).

Can a Mac run Gemma 4 26B-A4B?

Yes — Apple Silicon with 32 GB of unified memory or more (macOS lets the GPU use ~75% of it, ~24 GB) runs Gemma 4 26B-A4B at Q4_K_M.

Related

VRAM calculator for any model ·Token counter

Last updated 2026-08-03. Architecture figures from the model's published config.json; see themethodology.