RunsLocalGPU checkerLocal AI toolsv1.8 · data Oct 14, 2026

All models / Gemma 4 26B A4B

What GPU do I need to run Gemma 4 26B A4B?

Gemma 4 26B A4B has 25.2B parameters (3.8B active per token, mixture-of-experts). At Q4_K_M with 8K context it needs about 16.4 GB of memory. The smallest graphics card that runs it fully on the GPU is about 20 GB, such as the RX 7900 XT. Because only 3.8B parameters are active, it still runs usably with offload to system RAM on cards like the RTX 5080.

Check your own hardware

Memory by quantization (8K context)

QuantBits per weightMemory needed
Q8_08.527.9 GB
Q6_K6.5621.8 GB
Q5_K_M5.6919.0 GB
Q4_K_M4.8516.4 GB
Q3_K_M3.9113.4 GB
IQ2_M2.79.6 GB

Hardware that runs it fully in fast memory

HardwareMemoryBest quantEst. speed
RX 7900 XT20 GBQ5_K_M150+ tok/s
RTX 409024 GBQ6_K150+ tok/s
RTX 3090 Ti24 GBQ6_K150+ tok/s
RX 7900 XTX24 GBQ6_K150+ tok/s
RTX 309024 GBQ6_K150+ tok/s
M4 Pro 24GB24 GBQ4_K_M71 tok/s
RTX 509032 GBQ8_0150+ tok/s
M5 32GB32 GBQ6_K29 tok/s
M4 32GB32 GBQ6_K23 tok/s
M4 Max 36GB36 GBQ6_K79 tok/s
M4 Pro 48GB48 GBQ8_041 tok/s
M4 Max 64GB64 GBQ8_081 tok/s
Ryzen AI Max+ 395 64GB64 GBQ8_038 tok/s
RTX PRO 6000 Blackwell96 GBQ8_0150+ tok/s
M3 Ultra 96GB96 GBQ8_0122 tok/s
M4 Max 128GB128 GBQ8_081 tok/s
DGX Spark 128GB128 GBQ8_041 tok/s
Ryzen AI Max+ 395 128GB128 GBQ8_038 tok/s
M3 Ultra 256GB256 GBQ8_0122 tok/s
M3 Ultra 512GB512 GBQ8_0122 tok/s

Runs with offload to system RAM (slower)

Assumes 32 GB of DDR5 system RAM for offload. Estimates only; real speeds usually land within about 20%.