Can Molmo2 4B run on the L40S?
Yes. At 8K context the best fit is F16, using about 10.0 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~98 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|
| 2-bit (Q2_K) | 2.9 GB Fits | 3.5 GB Fits | 6.9 GB Fits | 475 |
| 3-bit (Q3_K_M) | 3.5 GB Fits | 4.0 GB Fits | 7.4 GB Fits | 364 |
| 4-bit (Q4_K_M) • | 3.8 GB Fits | 4.3 GB Fits | 7.7 GB Fits | 318 |
| 5-bit (Q5_K_M) | 4.5 GB Fits | 5.0 GB Fits | 8.4 GB Fits | 250 |
| 6-bit (Q6_K) | 5.0 GB Fits | 5.5 GB Fits | 8.9 GB Fits | 217 |
| 8-bit (Q8_0) • | 5.6 GB Fits | 6.2 GB Fits | 9.6 GB Fits | 184 |
| 16-bit (F16/BF16) • | 9.5 GB Fits | 10.0 GB Fits | 13.4 GB Fits | 98 |
• real GGUF file size from reubk/Molmo2-4B-GGUF. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.
More
Molmo2 4B 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 (allenai/Molmo2-4B), NVIDIA specs. Data as of 2026-11-17.