RunsOnMy ModelsGPUs & MacsBy VRAMBest GPUCan I run…?OllamaStart hereAPI pricesVRAM calculatorNew

Can Qwen2.5 VL 3B Instruct run on the RTX 3060 Ti?

Yes. At 8K context the best fit is Q8_0, using about 4.0 GB of the 7.6 GB usable on the RTX 3060 Ti. Estimated speed ceiling: ~136 tokens/sec.

+ 1.2 GB with image input. The figures on this page are for text only. With image input enabled, the vision projector (mmproj) file loads alongside the model: Q8_0 at 8K context then needs about 5.3 GB (still fits). llama.cpp's -hf downloads and loads it automatically unless you add --no-mmproj. Calculator with image input on.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K)2.1 GB Fits2.3 GB Fits3.1 GB Fits318
3-bit (Q3_K_M)2.5 GB Fits2.7 GB Fits3.5 GB Fits244
4-bit (Q4_K_M) •2.6 GB Fits2.8 GB Fits3.6 GB Fits232
5-bit (Q5_K_M)3.3 GB Fits3.5 GB Fits4.3 GB Fits167
6-bit (Q6_K) •3.2 GB Fits3.3 GB Fits4.2 GB Fits177
8-bit (Q8_0) •3.9 GB Fits4.0 GB Fits4.9 GB Fits136
16-bit (F16/BF16)7.8 GB No8.0 GB No8.8 GB No—

• real GGUF file size from lmstudio-community/Qwen2.5-VL-3B-Instruct-GGUF. Speed is a bandwidth ceiling (448 GB/s), not a benchmark. Methodology.

What fits instead

"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.

More

Qwen2.5 VL 3B Instruct on every GPU and Mac · Every model that runs on the RTX 3060 Ti · Other models that fit 8 GB VRAM · Try other settings in the calculator

Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs

Data: Hugging Face Hub (Qwen/Qwen2.5-VL-3B-Instruct), NVIDIA specs. Model data last checked 2026-11-18; refreshed hourly.