RunsOnMy ModelsGPUs & MacsBy VRAMBest GPUCan I run…?OllamaStart hereAPI pricesVRAM calculatorNew

Can GLM 4.5 Air run on the RTX 3090 Ti?

Not fully in VRAM. Even at 2-bit with 8K context, GLM 4.5 Air needs about 45.1 GB, more than the 22.8 GB usable on the RTX 3090 Ti. It can still run with offload (slower): Q4_K_M with part of the model in system RAM, about 6.5–13 tokens/sec (estimate). Details and command.

Runs with offload (slower)

Runs with offload 4-bit (Q4_K_M) at 8K context: all attention layers plus the experts of 10 of 46 MoE layers on the RTX 3090 Ti (about 20.5 GB), and the experts of the other 36 layers in system RAM (about 49.4 GB). Needs a PC with at least 64 GB of system RAM. Estimated speed: 6.5–13 tokens/sec (slow but usable), versus a bandwidth ceiling of ~101 tok/s if it fit fully in VRAM.

llama-server -hf bartowski/zai-org_GLM-4.5-Air-GGUF:Q4_K_M -c 8192 -ngl 999 --n-cpu-moe 36

Estimate, not a benchmark. The command downloads the GGUF from Hugging Face. The speed range spans dual-channel DDR4-3200 to DDR5-6000 system RAM; your CPU, RAM speed and llama.cpp version move it. If llama.cpp runs out of GPU memory, raise --n-cpu-moe by 1-2; if VRAM is left over, lower it for more speed. Recent llama.cpp builds can also pick the split automatically. How we estimate offload.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •44.3 GB No45.1 GB No49.4 GB No66.6 GB No—
3-bit (Q3_K_M) •53.1 GB No53.8 GB No58.1 GB No75.4 GB No—
4-bit (Q4_K_M) •69.9 GB No70.6 GB No74.9 GB No92.1 GB No—
5-bit (Q5_K_M) •79.4 GB No80.1 GB No84.4 GB No102 GB No—
6-bit (Q6_K) •93.8 GB No94.5 GB No98.8 GB No116 GB No—
8-bit (Q8_0) •111 GB No112 GB No116 GB No133 GB No—
16-bit (F16/BF16)207 GB No208 GB No212 GB No229 GB No—

• real GGUF file size from bartowski/zai-org_GLM-4.5-Air-GGUF. Speed is a bandwidth ceiling (1008 GB/s), not a benchmark. Methodology.

Hardware that does fit it

RTX A6000RTX 6000 AdaL40SMac M4 Pro 64GBMac M1 Max 64GBMac M4 Max 64GB

What fits instead

"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.

More

GLM 4.5 Air on every GPU and Mac · Every model that runs on the RTX 3090 Ti · Other models that fit 24 GB VRAM · Try other settings in the calculator

Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs

Data: Hugging Face Hub (zai-org/GLM-4.5-Air), NVIDIA specs. Model data last checked 2026-11-18; refreshed hourly.