Can deepseek vl2 tiny run on the L40S?
Yes. At 8K context the best fit is F16, using about 7.4 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~368 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | Max tok/s (est., 8K) |
|---|---|---|
| 2-bit (Q2_K) | 2.1 GB Fits | 1960 |
| 3-bit (Q3_K_M) | 2.4 GB Fits | 1504 |
| 4-bit (Q4_K_M) | 2.8 GB Fits | 1203 |
| 5-bit (Q5_K_M) | 3.1 GB Fits | 1032 |
| 6-bit (Q6_K) | 3.5 GB Fits | 896 |
| 8-bit (Q8_0) | 4.2 GB Fits | 692 |
| 16-bit (F16/BF16) | 7.2 GB Fits | 368 |
• real GGUF file size. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.
More
deepseek vl2 tiny 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 (deepseek-ai/deepseek-vl2-tiny), NVIDIA specs. Data as of 2026-11-17.