Can SmolVLM2 500M Video Instruct run on the RTX 4060?
Yes. At 8K context the best fit is F16, using about 1.6 GB of the 7.6 GB usable on the RTX 4060. Estimated speed ceiling: ~332 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | Max tok/s (est., 8K) |
|---|---|---|---|
| 2-bit (Q2_K) | 0.8 GB Fits | 1.0 GB Fits | 1429 |
| 3-bit (Q3_K_M) | 0.9 GB Fits | 1.0 GB Fits | 1097 |
| 4-bit (Q4_K_M) | 0.9 GB Fits | 1.1 GB Fits | 877 |
| 5-bit (Q5_K_M) | 1.0 GB Fits | 1.1 GB Fits | 752 |
| 6-bit (Q6_K) | 1.0 GB Fits | 1.2 GB Fits | 654 |
| 8-bit (Q8_0) • | 1.1 GB Fits | 1.2 GB Fits | 623 |
| 16-bit (F16/BF16) • | 1.4 GB Fits | 1.6 GB Fits | 332 |
• real GGUF file size from ggml-org/SmolVLM2-500M-Video-Instruct-GGUF. Speed is a bandwidth ceiling (272 GB/s), not a benchmark. Methodology.
More
SmolVLM2 500M Video Instruct on every GPU and Mac · Every model that runs on the RTX 4060 · Other models that fit 8 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (HuggingFaceTB/SmolVLM2-500M-Video-Instruct), NVIDIA specs. Data as of 2026-11-17.