Q4_K_M · 4K context
Gemma 4 26B-A4B (MoE) MoE model with 25.2B total parameters and 3.8B active per token · estimated Q4_K_M memory about 16GB
This is an MoE model in which 3.8B of the total 25.2B parameters are activated per token. Supports 256K (262,144) native contexts and provides fast decoding per load.
- Estimated Q4_K_M memory
- about 16GB
- total parameters
- 25.2B
- active parameter
- 3.8B
- native context
- 256K
Start with a model
Find hardware for this model
Also running document search?
Reserve this for embedding and reranking models sharing the same GPU or unified memory. Leave it at none if they run separately on the CPU.
Matching devices 55· 5 shown
Comfortable fit · Q4_K_M · Headroom 4.4GB
₩2,179,000
13.1 ~ 34.7 tok/s
Comfortable fit · Q4_K_M · Headroom 4.4GB
₩2,650,000
23.4 ~ 55.8 tok/s
Comfortable fit · Q4_K_M · Headroom 11.4GB
₩2,774,000
13.1 ~ 34.7 tok/s
Comfortable fit · Q4_K_M · Headroom 4.4GB
₩2,849,500
11.2 ~ 31 tok/s
Comfortable fit · Q4_K_M · Headroom 14.4GB
₩2,850,000
11.2 ~ 31 tok/s
Price is the range midpoint, with a basic host added for GPU-only prices. Speed is an estimate for the same model and 4K input.
Gemma 4 26B-A4B (MoE) Local Serving Run Recipe
Selecting a device changes the quantization, safety context, GPU offload, Flash Attention, KV cache, and batch start values.
If you want to choose equipment again
You can first narrow down the model configurations based on budget, noise level, operating system and whether it's acceptable used.