Can GLM 4.5 Air run on the A100 40GB?
Not fully in VRAM. Even at 2-bit with 8K context, GLM 4.5 Air needs about 45.1 GB, more than the 38.0 GB usable on the A100 40GB. It can still run with offload (slower): Q4_K_M with part of the model in system RAM, about 9.3–18 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 21 of 46 MoE layers on the A100 40GB (about 35.6 GB), and the experts of the other 25 layers in system RAM (about 34.3 GB). Needs a PC with at least 64 GB of system RAM. Estimated speed: 9.3–18 tokens/sec (usable for chat), versus a bandwidth ceiling of ~156 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 25
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
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | 44.3 GB No | 45.1 GB No | 49.4 GB No | 66.6 GB No | — |
| 3-bit (Q3_K_M) • | 53.1 GB No | 53.8 GB No | 58.1 GB No | 75.4 GB No | — |
| 4-bit (Q4_K_M) • | 69.9 GB No | 70.6 GB No | 74.9 GB No | 92.1 GB No | — |
| 5-bit (Q5_K_M) • | 79.4 GB No | 80.1 GB No | 84.4 GB No | 102 GB No | — |
| 6-bit (Q6_K) • | 93.8 GB No | 94.5 GB No | 98.8 GB No | 116 GB No | — |
| 8-bit (Q8_0) • | 111 GB No | 112 GB No | 116 GB No | 133 GB No | — |
| 16-bit (F16/BF16) | 207 GB No | 208 GB No | 212 GB No | 229 GB No | — |
• real GGUF file size from bartowski/zai-org_GLM-4.5-Air-GGUF. Speed is a bandwidth ceiling (1555 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
- Cheapest listed GPU that runs GLM 4.5 Air: DGX Spark (128 GB, launch MSRP $3,999) at Q6_K, fully in VRAM at 8K context.
- Largest popular model that fits the A100 40GB: Qwen3.6 35B A3B (36B, MoE) at Q6_K, about 29.9 GB at 8K context.
"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 A100 40GB · Other models that fit 48 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.