Will this model run on your hardware?
Can I run Google: Gemma 4 26B A4B locally?
- 26B parameters (4.0B active, MoE)
- context up to 256K
- licence: apache-2.0
- Weights: google/gemma-4-26B-A4B-it · GGUF: unsloth/gemma-4-26B-A4B-it-GGUF
Open this page in a browser to check it against your own GPU or Mac: the tool detects your hardware locally and shows the best quantization that fits, the memory it needs and an estimated speed.
Quantizations and file sizes
| Quantization | GGUF file |
|---|---|
| UD-Q6_K | 23.2 GB |
| UD-Q5_K_M | 21.2 GB |
| UD-Q4_K_M | 16.9 GB |
| MXFP4 | 16.6 GB |
| UD-Q3_K_M | 12.7 GB |
Run it locally
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q4_K_M
llama serve -hf unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q4_K_M
LM Studio: open unsloth/gemma-4-26B-A4B-it-GGUF