RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can SmolVLM2 500M Video Instruct run on the RTX 3080 12GB?

Yes. At 8K context the best fit is F16, using about 1.6 GB of the 11.4 GB usable on the RTX 3080 12GB. Estimated speed ceiling: ~1112 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctxMax tok/s (est., 8K)
2-bit (Q2_K)0.8 GB Fits1.0 GB Fits4792
3-bit (Q3_K_M)0.9 GB Fits1.0 GB Fits3677
4-bit (Q4_K_M)0.9 GB Fits1.1 GB Fits2940
5-bit (Q5_K_M)1.0 GB Fits1.1 GB Fits2522
6-bit (Q6_K)1.0 GB Fits1.2 GB Fits2192
8-bit (Q8_0) •1.1 GB Fits1.2 GB Fits2088
16-bit (F16/BF16) •1.4 GB Fits1.6 GB Fits1112

• real GGUF file size from ggml-org/SmolVLM2-500M-Video-Instruct-GGUF. Speed is a bandwidth ceiling (912 GB/s), not a benchmark. Methodology.

More

SmolVLM2 500M Video Instruct on every GPU and Mac · Every model that runs on the RTX 3080 12GB · Other models that fit 12 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (HuggingFaceTB/SmolVLM2-500M-Video-Instruct), NVIDIA specs. Model data last checked 2026-11-17; refreshed hourly.