Q4_K_M · 4K context
Qwen3.6 35B-A3B (MoE) MoE model with 35B total parameters and 3B active per token · estimated Q4_K_M memory about 22GB
This is an MoE model in which approximately 3B of the total 35B parameters are activated per token. Memory loading requires 35B capacity, and decoding bandwidth is consumed based on 3B active parameters.
- Estimated Q4_K_M memory
- about 22GB
- total parameters
- 35B
- active parameter
- 3B
- 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 51· 5 shown
Comfortable fit · Q4_K_M · Headroom 19.7GB
₩3,035,000
19.6 ~ 46.1 tok/s
Comfortable fit · Q4_K_M · Headroom 19.7GB
₩4,605,000
23.5 ~ 56.3 tok/s
Comfortable fit · Q4_K_M · Headroom 34.7GB
₩5,091,500
39.2 ~ 81.9 tok/s
Comfortable fit · Q4_K_M · Headroom 34.7GB
₩5,160,000
19.6 ~ 46.1 tok/s
Comfortable fit · Q4_K_M · Headroom 19.7GB
₩5,194,420
23.5 ~ 56.3 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.
Qwen3.6 35B-A3B (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.