RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can Qwen2.5 VL 3B Instruct run on the RTX 4090?

Yes. At 8K context the best fit is F16, using about 6.7 GB of the 22.8 GB usable on the RTX 4090. Estimated speed ceiling: ~163 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •2.0 GB Fits2.2 GB Fits3.0 GB Fits791
3-bit (Q3_K_M) •2.3 GB Fits2.5 GB Fits3.3 GB Fits634
4-bit (Q4_K_M) •2.6 GB Fits2.8 GB Fits3.6 GB Fits522
5-bit (Q5_K_M) •2.9 GB Fits3.0 GB Fits3.9 GB Fits453
6-bit (Q6_K) •3.2 GB Fits3.3 GB Fits4.2 GB Fits397
8-bit (Q8_0) •3.9 GB Fits4.0 GB Fits4.9 GB Fits307
16-bit (F16/BF16) •6.6 GB Fits6.7 GB Fits7.6 GB Fits163

• real GGUF file size from unsloth/Qwen2.5-VL-3B-Instruct-GGUF. Speed is a bandwidth ceiling (1008 GB/s), not a benchmark. Methodology.

More

Qwen2.5 VL 3B Instruct on every GPU and Mac · Every model that runs on the RTX 4090 · Other models that fit 24 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (Qwen/Qwen2.5-VL-3B-Instruct), NVIDIA specs. Data as of 2026-11-17.