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

Can Llama 3.2 3B run on the A100 40GB?

Yes. At 8K context the best fit is F16, using about 7.5 GB of the 38.0 GB usable on the A100 40GB. Estimated speed ceiling: ~242 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •2.5 GB Fits2.9 GB Fits5.5 GB Fits16.0 GB Fits1041
3-bit (Q3_K_M) •2.8 GB Fits3.2 GB Fits5.9 GB Fits16.4 GB Fits838
4-bit (Q4_K_M) •3.2 GB Fits3.6 GB Fits6.2 GB Fits16.7 GB Fits694
5-bit (Q5_K_M) •3.5 GB Fits3.9 GB Fits6.6 GB Fits17.1 GB Fits600
6-bit (Q6_K) •3.8 GB Fits4.3 GB Fits6.9 GB Fits17.4 GB Fits524
8-bit (Q8_0) •4.7 GB Fits5.1 GB Fits7.7 GB Fits18.2 GB Fits405
16-bit (F16/BF16)7.1 GB Fits7.5 GB Fits10.1 GB Fits20.6 GB Fits242

• real GGUF file size from QuantFactory/OneLLM-Doey-V1-Llama-3.2-3B-GGUF. Speed is a bandwidth ceiling (1555 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

Llama 3.2 3B on every GPU and Mac · Every model that runs on the A100 40GB · Other models that fit 48 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 (meta-llama/Llama-3.2-3B), NVIDIA specs. Model data last checked 2026-11-17; refreshed hourly.