Can GLM 4.7 Flash run on the L40S?
Yes. At 8K context the best fit is Q8_0, using about 30.8 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~184 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) • | 11.5 GB Fits | 11.7 GB Fits | 12.9 GB Fits | 17.9 GB Fits | 516 |
| 3-bit (Q3_K_M) • | 14.5 GB Fits | 14.7 GB Fits | 16.0 GB Fits | 20.9 GB Fits | 401 |
| 4-bit (Q4_K_M) • | 18.0 GB Fits | 18.2 GB Fits | 19.4 GB Fits | 24.4 GB Fits | 320 |
| 5-bit (Q5_K_M) • | 20.8 GB Fits | 21.0 GB Fits | 22.3 GB Fits | 27.2 GB Fits | 273 |
| 6-bit (Q6_K) • | 23.9 GB Fits | 24.1 GB Fits | 25.3 GB Fits | 30.3 GB Fits | 237 |
| 8-bit (Q8_0) • | 30.6 GB Fits | 30.8 GB Fits | 32.0 GB Fits | 37.0 GB Fits | 184 |
| 16-bit (F16/BF16) • | 56.7 GB No | 56.9 GB No | 58.1 GB No | 63.1 GB No | — |
• real GGUF file size from unsloth/GLM-4.7-Flash-GGUF. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.
More
GLM 4.7 Flash on every GPU and Mac · Every model that runs on the L40S · Other models that fit 48 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (zai-org/GLM-4.7-Flash), NVIDIA specs. Data as of 2026-11-17.