RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can medgemma 4b it run on the L40S?

Yes. At 8K context the best fit is F16, using about 8.4 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~111 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •2.7 GB Fits2.8 GB Fits3.2 GB Fits5.1 GB Fits500
3-bit (Q3_K_M) •3.0 GB Fits3.1 GB Fits3.6 GB Fits5.5 GB Fits412
4-bit (Q4_K_M) •3.4 GB Fits3.5 GB Fits4.0 GB Fits5.8 GB Fits347
5-bit (Q5_K_M) •3.7 GB Fits3.8 GB Fits4.3 GB Fits6.1 GB Fits305
6-bit (Q6_K) •4.1 GB Fits4.1 GB Fits4.6 GB Fits6.5 GB Fits271
8-bit (Q8_0) •4.9 GB Fits5.0 GB Fits5.5 GB Fits7.4 GB Fits209
16-bit (F16/BF16) •8.3 GB Fits8.4 GB Fits8.9 GB Fits10.7 GB Fits111

• real GGUF file size from unsloth/medgemma-4b-it-GGUF. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.

More

medgemma 4b it 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 (google/medgemma-4b-it), NVIDIA specs. Data as of 2026-11-17.